вход по аккаунту


The trabecular anatomy of late stages of experimental dietary cirrhosis. Its pathogenesis in terms of rappaport's structural unit

код для вставкиСкачать
l s 2
Banting and Best Department of Medical Research, University of Toronto,
Toronto, Ontario, Canada
Rappaport, Borowy, Lougheed and Lotto have presented
evidence which indicates that the smallest functional subdivision of the liver is a small piriform mass of parenchyma
enveloping the terminal branches of the portal vein, hepatic
artery and biliary tree. This structural urzit comprises portions of all those hexagonal lobules (as conventionally conceived) which lie along the course of any one terminal portal
venule. Rappaport’s structural unit is made up of adjacent
portions of several conventional lobules, and is smaller than
a single lobule. One important implication of this concept
is that blood leaves venules and enters the sinusoidal plexus
in regions that are not always immediately periportal in the
conventional sense. This reorientation of blood supply in
relation to hepatic architecture has enabled us to comprehend
the tridimensional system adhered to by the fibrous trabeculae
a t all stages in their development in experimental dietary
‘This work was supported by a grant from the National Cancer Iustitute of
Canada to Professors C. H. Best, E. A. Sellers, and t o the author.
* A synopsis of this work was presented a t the Eleventh Conference on Lirer
I n j u r y spoilsored by the Josiali Macy, Jr. Foundation, New York City, May 30,
3T1iis structure is described in the preceding paper i n this issue, entitled
Subdivision of hexagonal liver lobules into a structural and functional unit.
Role i n liver physiology and pathology,” by A. M. Rappaport, E. J. Borowy,
W. M. Lougheed and W. N. Lotto.
cirrhosis. I n terms of the classical concepts of hepatic microanatomy, the spatial patterns of the trabeculae had previously appeared incomprehensible and even paradoxical. The
purpose of this paper is to present a new interpretation of
the late stages of cirrhosis in choline-deficient rats. Since
this type of liver damage is closely related to that frequently
associated with chronic alcoholism in man (Best et al., '49),
the results may prove of interest to clinical, as well as to
experimental, pa thologis t s.
I n order to state more clearly some of the problems involved, it is necessary to summarize briefly the results of
previous studies of this type of cirrhosis (Hartroft, '50;
Hartroft and Ridout, '51; Hartroft and Sellers, '52). At
every stage the sites most severely affected are always
farthest distant from the precapillary terminations of the
portal vein and hepatic artery. The course of the pathogenesis of the cirrhosis is initiated by the intracellular
accumulation of abnormal amounts of fat, followed by an
extracellular storage phase within fatty cysts. The first cysts
are formed whenever adjacent parenchymal cells become so
distended by fat that their limiting membranes are compressed
together and transformed into a single, tenuous septum which
is eventually torn. The result is release of the fat globules
from the cells so that they pass from an intracellular to an
extracellular phase where they fuse into a single mass, still
enclosed, however, by the group of two or more cells from
which they have escaped. The latter now form the wall of
a small cyst filled with fat. Adjacent fatty cysts may in
their turn rupture and fuse to form larger ones. Rupture of
very large cysts may tear adjacent bile canaliculi or blood
sinusoids, permitting escape of liberated fat into either the
biliary or the vascular system. The parenchymal and stromal
remnants of torn cysts are compressed by the surrounding
parenchyma, and with addition of more and more cyst-remnants, trabeculae are evolved. I n single sections of livers
in advanced but actively progressive cirrhosis, every stage
may sometimes be encountered within the same structural
?he lesions (intracellular f a t ; extracellular fat within cysts;
rupture of cysts and compression of their remnants into
trabeculae) always make their initial appearance in regions
immediately surrounding central veins. These sites are farthest distant from the junctions of sinusoids with afferent
venules or arterioles. Areas affected subsequently are those
that link together adjacent central veins, for, second to the
pericentral zones, they are farthest from the points at which
fresh blood enters the sinusoidal plexus. Chains of fatty
cysts form annular patterns foreshadowing the distribution
of the fibrosis which follows. The trabeculae enclose nodules
of less severely affected parenchyma in which may be found
relatively large branches of the portal vein and hepatic artery.
Up to this point, the progress of the lesions is easily interpreted in terms of the conventional hexagonal lobule. Investigators who have studied the early stages of this type
of cirrhosis (Lillie et al., '42; Glynn et al., '48; Hartroft, '49)
agree that the fibrosis is clearly of nonportal distribution.
I n late stages, however, an apparently anomalous turn is
taken by the lesions. Fatty cysts or their remnants, compressed into slender trabeculae, extend from the annular
bands toward the enclosed portal triad (fig. 3). These errant
trabeculae appear to pierce the substance of the conventional
lobule without any apparent relation to recognizable landmarks or to the course of the blood flow. It is only in the light
of Rappaport's concept of the hepatic structural unit that
this last development can be explained. Eventually these
anomalous offshoots link portal and central veins and appear
to divide haphazardly the conventional lobule (figs. 4, 5 ) . I n
this way large portal triads become surrounded by scar tissue,
and it is not surprising that initial observers of this type
of experimental cirrhosis classified it as portal (Gyorgy and
Goldblatt, '39), for they were studying the late stages of the
lesion. Not until subsequent investigators (loc. cit.) had ex-
amined the early stages was the initial nonportal distribution
of the fibrosis recognized.
It appears that at a certain stage fibrosis, which up to this
period had been of a clearly nonportal origin, spreads through
the liver lobule to enclose portal triads, despite the fact that
at every previous phase in the development of the lesions,
parenchyma closest to portal areas had been affected least.
An attempt has been made to resolve this apparently paradoxical alteration of the nature of the late stages of the
lesion-i.e., from nonportal to periportal. With the aid of
special preparations of livers of rats sacrificed at frequent
intervals during an experimental period extending up t o
16 months, the tridimensional course of the trabecular pattern has been studied. The results will be interpreted in the
light of Rappaport's concept of the structural unit, for only
when viewed in this manner can the course of the advanced
stages of cirrhosis be comprehended.
Microsections and special preparations from over 350 rats
of the Wistar strain have been studied in the course of this
investigation. From the time they were weanlings, 200 of
these (equal distribution of sexes) received a choline-deficient
diet similar to that described by Copeland and Salmon ('46).
The composition of this diet and details concerning its preparation and the care of the animals have been published
(Hartroft and Ridout, '51). Forty male weanling rats were
fed a synthetic diet of a similar nature, with gelatin, casein,
and arachine in proportions previously reported (Hartrof t
and Ridout, '51). I n addition, 112 male rats at the age of
78-80 days (125-150gm) were fed ad libitum a diet more
severely hypolipotropic than either of the others. This diet
had the following percentage composition : peanut meal (solvent process) 6 ; alpha soya protein (Glidden's) 6 ; hydrogenated vegetable fat (Primex) 20; salts 3 (Hartroft and
Ridout, '51) ; vitamin mixture 1 (Hartroft and Ridout, '51) ;
cellu flour 2 ; sucrose 61.8; cystine 0.15; alpha tocopheryl
acetate 0.035; and cod liver oil concentrate 0.015. Animals
were sacrificed at frequent intervals throughout all these
dietary regimens, and some were maintained as long as 16
months. Animals on the very severely hypolipotropic diet
were not followed for a period longer than 6 months.
From all animals, frozen sections of liver (formol-calcium
fixation) were stained to demonstrate fat by Wilson’s technique (Wilson, ’50), and paraffin sections of duplicate blocks
(Bouin-fixed) were stained with hematoxylin and eosin. I n
addition, from representative animals in various stages in
the development of the cirrhosis, special preparations for
tridimensional studies of the liver were made as follows.
Under ether anesthesia the animal’s heart was exposed and
a plastic cannula was introduced into the tip of the left ventricle via a small stab wound, and ligatured in place. The
cannula was attached by rubber tubing to a Y-tube connected
with two reservoirs. One contained an isotonic flushing solution of 0.4% NaCl and 0.5% NaNO,, designed to produce
maximum dilation of all vessels. The other was filled with
one part India ink (Higgins) and two parts distilled water
to which a few drops of ammonia had been added, and the
whole filtered. When the cardiac cannula was in position,
the flushing solution was introduced intravenously under a
pressure of 100mm Hg. A pathway for escape of fluid was
provided by severing the thoracic portion of the inferior vena
cava, after ligaturing the portion nearest the heart to prevent
escape of fluid from the heart by retrograde flow. The vascular system was flushed until the outflow was colorless or nearly
so, The diluted, filtered ink was then introduced and continued
under the same pressure until the animal’s skin became dark
grey. A practical guide for judging the completeness of the
injection was afforded by the mesentery. By exposing a
small portion of it through an abdominal incision, stages in
the injection could be directly observed with a small handlens. At first the larger vessels became visible as black lines,
and later others, probably of arteriolar o r venular size, be-
came manifest. When the injection was complete the
whole mesenteric web was uniformly black at this low
magnification, and probably indicated complete filling of
the dilated capillary bed. When this stage was reached,
inflow of ink was continued only one or two minutes longer.
Tissues and blocks of viscera were fixed by immersion
in formol-saline and slices ranging from 100 to 200 p in thickness were cut on the freezing microtome. The latter were
dehydrated in ascending strengths of isopropyl alcohol, passed
through benzene, and finally cleared in benzyl benzoate. When
observed with the aid of the binocular, dissecting microscope,
they afforded three-dimensional visualization. Serial slices
of this type permitted rapid reconstruction and spatial observation of more extensive portions of the vascular tree.
I n addition to the thick slices, thin ( 3 p ) frozen and paraffin
sections were prepared as described above, since the presence
of the injection mass in the vessels did not interfere with
the usual sectioning and staining techniques. A variety of
special stains to demonstrate connective tissue, glycogen,
ceroid, etc., were applied to selected sections. I n addition,
a group of 10 normal animals were injected with ink, and
thick slices of the fixed liver tissue cleared in benzyl benzoate.
To 5 rats in a special group a single oral dose of 0.2 em3 of
a 20% solution of carbon tetrachloride in corn oil was administered 24 hours before they were injected intravenously
with ink and sacrificed. The object of this procedure is explained below.
T h e functional versus the purely anatomical concept of the
term, periportal. All parenchyma that adjoins portal canals
is not necessarily periportal in the connotation that blood
supplies it directly from the portal vein rather than from
other sinusoids situated more proximally. This fact is well
demonstrated in the thick cleared slices of livers of inkinjected rats which had received a small amount of carbon
tetrachloride 24 hours before they were sacrificed. I n these
preparations, the hepatotoxin has served to label the centro-
lobular regions, because parenchyma adjacent to central veins
can be identified by sinusoids that have been compressed by
swollen liver cells (fig. 1). Conversely, veins that are surrounded by sinusoids of normal diameter are thereby recognized as portal, for small doses of carbon tetrachloride do
not produce morphologic changes at these sites. By studying
these preparations, it has been possible to demonstrate characteristics of the terminal arborizations of the portal venous
tree which serve to identify them in similar liver slices of
animals that have not been given carbon tetrachloride.
As is the case of afferent channels elsewhere, few vessels
of capillary size communicate directly with branches of the
portal vein larger than the order of venules. As in other
organs, afferent vessels larger than the latter function merely
to conduct blood to the distributing venules (fig. 2). I n ordinary microsections, conducting veins (with their accompanying artery and bile duct) are prominent, easily recognized
landmarks. Small distributing venules, however, which are
of greater importance as landmarks of functional boundaries,
are inconspicuous and can be found only under the higher
magnifications of the microscope. Furthermore, because venules are short and narrow, they occur only in one or two
sections, whereas larger branches are of sufficient length
and diameter that portions of each will be found in several.
I n thick (100-200 v) slices of injected livers, these factors,
which obscure the general vascular pattern, are minimized.
Narrow collars of non-injected tissue frequently can be
observed around conducting veins, indicating that they do
not communicate directly with adjacent sinusoids (fig. 2).
When apparent exceptions to this rule have been encountered,
further observation with the binocular dissecting microscope
focussed at varying levels within the slice, has revealed that
sinusoids which at first seemed to communicate directly with
large conducting veins in reality course over the latter. This
is demonstrated in figure 2, for in those places where ink
has dropped out of the conducting vein, it becomes obvious
that the sinusoidal plexus overlies the vessel. Had the in-
Section mass remained ia situ, it would have appeared as if
the sinusoids ran into, rather than over the vein. I n the same
way pseudo-communications of this type might seem to be
present in preparations used in the study of the hepatic circulation in vivo (quartz rod illumination and similar methods).
We have encountered few exceptions to the rule that sinusoids adjacent to large conducting veins do not receive their
blood directly from the latter, but instead from nearby terminal portal venules via other, intervening sinusoids. Portal
venules resemble arterioles generally, in that their ramifications with vessels of capillary diameter imitate the manner
in which an elm tree gives off its finest branches and twigs.
Terminal arborizations of the portal veins occur at gentle
angles which would not produce acute reversals in the direction of blood flow.
On the other hand, sinusoids that drain into central veins
(the latter are easily identified in figs. 1 and 2 by the pathologic compression of surrounding sinusoids secondary to
carbon tetrachloride administration) form converging patterns around the vessels. This characteristic serves to distinguish the latter from the tree-like terminations of the
portal vein in cleared slices of livers in which centrolobular
labelling has not been effected with carbon tetrachloride.
Direct communications between sinusoids and even relatively
large radicles of the hepatic vein appear to exist.
I n a geographic sense, parenchyma adjacent to large conducting branches of the portal vein is clearly periportal. But
from a physiologic standpoint, this juxtaposition appears
quite fortuitous and incidental, and does not distinguish parenchyma in these regions from that elsewhere in the unique
functional sense that characterizes juxtaposition with terminal venules. The latter mark the sites a t which fresh blood
enters the parenchyma, which here has access to blood that
has not previously been exposed t o other liver cells. But this
is not true for tissue adjacent to conducting veins, which is
periportal only in a spatial sense, whereas that around venules
is not only anatomically but also functionally portal. Cells
around large conducting veins are no more favored in the
latter way than are those in other regions of the liver lobule
which are not only functionally but also obviously anatomically nonportal. The importance of clearly recognizing this
distinction between geographic and physiologic types of periportal tissue becomes apparent when one attempts to comprehend the pathogenesis of the later stages of experimental
T h e consistently %onportal nature of the trabeculae ia
dietary cirrhosis. It appeared difficult to explain why in the
late stages of dietary cirrhosis the fibrosis apparently becomes periportal when previously it had followed a distribution characterized above all by its nonportal nature. Study
of thick slices of ink-injected livers from rats at various
stages of disease has resolved the apparent paradox. I n these
preparations, portal vessels can always be distinguished from
radicles of the hepatic vein by the characteristics described
in the previous section. The regions of fibrosis are easily
identified because destruction of the sinusoidal pattern with
replacement by poorly vascularized tissue has rendered them
particularly prominent as non-injected clear zones.
Figure 3 shows in cross section a large conducting portal
vein from which radiate a number of venules terminating in
sinusoids. Surrounding this field, taken from a cholinedeficient rat in an early stage of dietary cirrhosis, is an
annulus of fibrous tissue which links adjacent central veins
in the manner previously described (Hartroft, '50). I n addition, trabeculae have extended toward the conducting portal
vein (center of field). The terminal venules of the portal
vein, which lie midway between these radical trabeculae, are
surrounded by piriform masses of intact parenchyma. T h e
cirrhosis at this stage has dissected Rappaport's structural
unit. Furthermore, it is apparent that in their course these
secondary radical trabeculae have adhered to the pathonomy
which holds f o r the distribution of all liver lesions in choline
deficiency ; i.e., nonportal regions are those most severely
affected. The trabeculae are midway between the terminal
portal venules, which indicate the positions of the functionally periporotal parenchyma that is still intact. I n one or
two places fibrosis has actually made contact with the adventitial sheath of the conducting vein, and this would seem
to constitute a contradiction of the nonportal nature of the
fibrosis, were it not for the fact that parenchyma in these
positions is periportal only in a fortuitous, geographic sense.
To reach the latter regions, the blood had to come from the
terminal venules via the adjacent sinusoids, for there are no
direct communications here with the large conducting vein.
These sites of fibrosis, adjacent to the conducting vein, are
just as functionally nonportal as others which are quite obviously nonportal, both anatomically and functionally.
I n similar preparations of later stages in the cirrhotic process, it can be shown that considerable portions of large conducting portal veins become surrounded by fibrous tissue by
a n extension of the process described above. Parenchyma
that had undergone fibrotic replacement in these positions
was always previously periportal in an anatomical sense only,
for it did not receive its blood supply directly from precapillary portions of the portal venous tree. Various stages in
the formation of geographic periportal fibrosis are shown
in figure 4, in which its development can be traced within a
single field. Two conducting portal veins are shown cut
lengthwise, and from these a number of terminal venules
arise. Each of the latter ends in a clump of sinusoids supplying a mass of parenchyma which has been reduced to a varying
degree by the cirrhotic process. This parenchymal mass
represents Rappaport’s berry-shaped structural unit, which,
in every case, has been most severely damaged at its periphery.
I n some instances, only the outermost nonportal zones have
been outlined by the fibrotic process (upper right), while in
other cases (lower left) entire acini have been destroyed but
for remnants supplied by groups of withered sinusoids immediately surrounding the terminal venules. Some of the
acinar peripheries lie next the conducting portal veins, and
despite this geographic periportal position they have shared
in the fibrotic replacement to the same extent as have other
similar portions of the acini which do not happen t o abut a
conducting vein. The result is a fibrosis which, although
anatomically periportal, is in fact nonportal in relation to
circulatory dynamics. It can be observed also that the entire
length of the corresponding radicles of the hepatic vein, which
lie between their portal counterparts, are completely encircled by fibrous tissue. This observation demonstrates the
essentially nonportal nature of the cirrhosis even at a stage
where appreciable portions of large branches of the portal
tree are embedded in scar tissue.
When the fibrotic replacement of physiologically nonportal,
but geographically periportal, parenchyma has advanced to
a later stage, every large vein, whether portal or central
(fig. 5 ) is embedded in fibrous tissue, for the central veins had
been so affected early in the course of the disease. Only the
terminal portal venules are still surrounded by intact sinusoids, indicating that the parenchyma here has been least
affected. These are the sites of the compensatory hyperplasia
which leads to the eventual formation of parenchymal nodules.
I n ordinary thin (3 p ) sections of cirrhotic livers (fig. 6)
the large conducting branches of the portal vein (with the
hepatic artery and bile duct) are exceedingly prominent, not
only because of their size but also because of the abundant
fibrosis which surrounds them. Conversely, the absence of
fibrous tissue around the terminal portal venules (figs. 6, 7),
and their small size, renders them so inconspicuous that they
may completely escape notice. They are frequently not even
included in the plane of the section because of their relatively
short length and narrow diameter, and therefore a number
of persisting parenchymal masses usually must be examined
before one is found. I n sections of cirrhotic livers from both
choline-deficient rats and from cases of alcoholism in man,
we have never encountered significant degrees of fibrosis
around portal venules, and in most instances the surrounding
parenchyma showed evidence of compensatory hypertrophy
rather than death and fibrotic replacement.
T h e significance of the term, periportal, i.n descriptions o f
pathological lesions. It has been emphasized above that the
term, periportal, may have either an anatomical or a structural definition. Although when applied to some regions (those
around terminal portal venules), it can be used justly with
both connotations, this does not hold when applied to sites
adjacent to large conducting trunks of the portal veins, where
it can have only an anatomical meaning. We do not believe
that investigators have generally recognized the need for
such a distinction when using the term in pathological descriptions, and this ambiguity may be responsible f o r some
of the confusion in our present concepts of cirrhosis.
The advantages to be gained by defining the exact distribution of a liver lesion lie in whatever clues to the nature of
the condition may thus be provided. For example, it is
reasonable to suspect that choline deficiency may least affect
cells which have first access to blood reaching the liver, because small amounts of lipotropic substances still present
in the blood would be available to them, perhaps, but not to
other cells to which the blood subsequently passes. The possible existence of such a pathonomy is obscured unless the
distribution of the lesions is clearly visualized in relation to
the blood supply and to the organization of the structural
units of the liver. With this example in mind, it is strongly
urged that the term periportal be used in its functional sense
only, and not to describe the result of mere spatial juxtaposition (which bears no relation to the organization of the
fundamental, structural units of the organ). No one would
attempt to describe the lobular distribution of a renal lesion
on the basis of its relation to arcuate vessels, for the latter
pass through renal lobules in an almost haphazard fashion.
The interlobular arterioles and venules are the reliable landmarks for lobular orientation in the kidney. The terminal
portal venules of the liver might be considered analogous
to the renal interlobular vessels, and like the latter, we believe
they are the only trustworthy guides to a morphologic classi-
fication of lesions that can be of any value in physiopathic
interpretations of hepatic disease.
The termiaal portal venule; its relation t o the convemtional
hepatic lobule and to Rappaport's structural unit. It has been
pointed out that this important vessel is small and inconspicuous not only in the normal liver, but also particularly in the
fatty or cirrhotic organ, for reasons already presented. Its
diameter is only slightly greater than that of a red blood cell,
and intracellular fat vacuoles, not to mention fatty cysts, niay
have diameters several times that of the venule. It is accompanied by a small bile ductule (fig. 7), but we have not
been able to identify a corresponding branch of the hepatic
artery in either ordinary sections or the thick, cleared slices.
For these reasons, the terminal portal venule obviously does
not correspond to that branch of the portal vein found in the
prominent type of portal triads usually used for the identification of the periphery of the conventional hexagonal lobule.
Branches in this position most frequently occupy a more
proximal position in the portal venous tree and are of the
conducting variety. What we have described as the terminal
portal venule has been previously regarded as a sublobular
vessel, which Elias ( '49a, '49b) on the basis of reconstructions
of serial sections has considered to be arterial in nature.
However, in our tridimensional slices, direct observation indicates its continuity with the portal venous tree (figs. 3 , 4 , 5 ) .
It should be emphasized that although the terminal portal
venule may be consistently visualized in the thick slices,
because of the prominence afforded it by the injection media,
it is so inconspicuous in ordinary sections that it may easily
escape detection unless a special search for it is made (fig. 7 ) .
The terminal portal venule is the chief afferent vessel of
Rappaport's structural unit, and frequently runs in the center
of the mass of berry-shaped parenchyma comprising this structural unit. Once illustrated in figures 3, 4 and 5, it becomes
comprehensible in terms of the blood supply of the structural
unit. Every stage of experimental dietary cirrhosis in cholinedeficient rats, includinp advanced cases, follows the patho-
nomic principle that cells farthest from the entry of fresh
blood into the parenchyma are most severely affected. The
severe fatty or fibrotic changes in parenchyma adjacent to
large conducting branches of the portal vein, when viewed in
this light, is a confirmation, rather than a contradiction of
the rule.
Criteria f o r t h e localization of hepatic lesions. The morbid
anatomist dealing with autopsy material cannot prepare thick,
cleared slices of injected livers as was done for this study.
The localization of lesions must be determined from an examination of routine micro-sections. The most prominent,
and easily identified, structures in thin sections are the large
portal triads, which contain a conducting branch of the portal
vein. A glance at figure 4 will reveal, however, that these
vessels do not bear any constant relation to the functional
unit of the liver, Rappaport's structural unit. Conducting
veins often pass through their peripheries, whereas at the
sites of origin of terminal portal venules the vessels lie near
their centers. Tissue surrounding these large portal triads
is more often nonportal in a functional sense than periportal.
I n thin sections central veins may be distinguished with difficulty from portal veins when the latter have been sectioned
in a plane that does not include accompanying bile ducts or
arteries. For these reasons, we believe the only reliable
landmark is the terminal portal venule accompanied by its
small bile ductule (fig. 7). The pathologist may have to
examine a relatively large portion of the section to locate
these vessels, but once they are found he is certain of his
bearings. All that is then necessary is to determine whether
the lesions in question are consistently present or absent
around the venules, and classify the distribution as functionally periportal o r nonportal on this basis.
T h e distribution of lesions in cirrhotic livers of cholinedeficievzt rats and chronic alcoholism in man. When the criteria
described above were applied to cases of alcoholic cirrhosis
in man, it was found that the terminal portal venules were
surrounded by parenchyma that had escaped fibrotic replace-
ment in every instance (figs. 6 , 7 ) . Frequently the surrounding
tissue had undergone hypertrophy of a compensatory nature.
I n these respects it resembles the distribution of lesions in
experimental dietary cirrhosis. The initial observers of the
latter condition (Gyorgy and Goldblatt, '39) also considered
its pathological anatomy to be similar to that of alcoholic
cirrhosis in man, for they observed the fibrosis around the
large conducting branches of the portal vein and described
the fibrosis as periportal. Alcoholic cirrhosis is the classical
example of the so-called portal type of cirrhosis, and we suggest that this is true only in a geographic sense as in the
choline-deficient rat. When the anatomy of both types of
cirrhosis is considered in the light of Rappaport's structural
unit, and the distribution of the lesions is studied in relation
to functionally periportal tissue, the evidence indicates that
morphologically the clinical condition has its experimental
counterpart in choline-deficient rats. Evidence has been published (Best et al., '49) which suggests that the etiology of
the two conditions may also have a common basis in a dietary
lack of lipotropic factors.
The distribution of fibrous tissue in the initial stages of
cirrhosis in choline-deficient rats has been recognized to be
clearly nonportal. I n later stages, trabeculae extend around
large branches of the portal vein in an apparently haphazard
fashion. Thick slices (100-200 p) of ink-injected livers were
prepared from rats sacrificed at intervals up to 16 months
from the time they had been put on one of several cholinedeficient diets. From observations of the lesions in these
special preparations, as well a s in ordinary micro-sections,
the pathogenesis of late stages of the cirrhosis was correlated
with the early stages.
Parenchyma adjacent to large conducting branches of the
portal vein, although periportal in a geographic sense, is
nonportal in the functional use of the term, because tissue
in these situations is not supplied by sinusoids which arise
directly from the vein. Much of the tissue around large
conducting vessels is supplied by blood that reaches it
from terminal portal venules, via sinusoidal pathways of
appreciable length.
From annular, nonportal trabeculae which link neighboring
central veins, fibrous bands extend toward large branches of
the portal vein, following pathways midway between terminal
portal venules, thereby replacing the functionally nonportal
tissue in these regions, By a continuation of this process, the
regions around conducting portal veins become fibrotic. When
this stage in the development of the cirrhosis has been reached,
although the fibrosis is geographically periportal, it can still
be regarded as functionally nonportal.
By trabecular isolation of the parenchymal masses around
terminal portal venules, Rappaport’s structural unit * is dissected and its boundaries are delineated. Without this concept of the organization of the hcpatic architecture, the
pattern assumed by the fibrosis would be difficult to comprehend.
The terminal portal venule is regarded as the essential
landmark for the lobular orientation of hepatic lesions. The
relations of this vessel and its morphological characteristics
are illustrated.
The cirrhosis associated with chronic alcoholism in man is
believed to be periportal in a geographic sense only, for the
terminal portal venules in these livers are nonfibrotic. I n
this respect, as well as others, it should probably be regarded
as the clinical counterpart of the cirrhosis induced in rats
by low-choline diets.
We wish to thank Mrs. Margaret Cornell, Librarian, Banting and Best Department of Medical Research, University of
Toronto, for her valuable assistance in the preparation of
this paper.
See footnote 3, page 71.
C. C. LUCASAND J. H. RIDOUT 1949 Liver damage
produced by feeding alcohol or sugar and its prevention by choline.
Brit. Med. J., 9: 1001-1006.
1946 The occurrence of neoplasms in
the liver, lungs and other tissues of rats as a result of prolonged
choline deficiency. Am. J. Path., ,92: 1058-1067.
ELIAS,H. 1949a A re-examination of the structure of the mammalian liver.
I. Parenchymal architecture. Am. J. Anat., 8 4 : 311-333.
194913 A re-examination of the structure of mammalian liver, hepatic lobule, and its relation to vascular and biliary systems. Am. J.
Anat., 85 : 379-456.
AND 0. LINDAN1948 Experimental production
and development of diffuse hepatic fibrosis ( ( (portal cirrhosis ”).
Brit. J. Exp. Path., 29: 1-10.
GYORGY,P., AND H. GOLDBLATT1939 Hepatic injury on a nutritional basis in
rats. J. Exp. Med., 7 0 : 185-192.
W. S. 1949 The locus of the beginning dietary cirrhosis. Transactions of the Eighth Conference on Liver Injury. Josiah Macy, JY.
Foundation, 126-155.
1950 Accumulation of f a t in liver cells and i n lipodiastaemata
preceding experimental dietary cirrhosis. Anat. Rec., 106 : 61-87.
w. AND J. H. RIDOUT 1951 Pathogenesis of the cirrhosis produced
by choline deficiency. Escape of lipid from f a t t y hepatic cysts into
the biliary and vascular systems. Am. J . Path., 27: 951-990.
HARTROFT,W. S., AXD E. A. SELLERS1952 The dissolution of f a t t y cysts in
precirrhotic and cirrhotic livers of eholine-deficient rats treated with
lipotropic factors. Am. J. Path., 28: 387-399.
F. S. DAFT A N D J. V. LOWRY1942
Histogenesis and repair of the hepatic cirrhosis in rats produced on
low protein diets and preventable with choline. U. S. Public Health
Rep., 5 7 : 502-504.
WILSON,W. 1950 A trichrome method f o r staining f a t with oil red 0 in frozen
sections. J. Tech. Methods, 31: 216-220.
1 Liver. The vessels, filled with injected ink, appear black, aiid the intcrveiling parenchynia IS almost iiivisible i n this thick (100 p ) slice which was cleared
iii bcnzgl benzoate. For technical details, see text. I n the center of the field is
i~ ressc.1 wliicli c a n be identified as a radicle of the hepatic vein by the surrounding
siiiusoitls which converge toward it from all directions. The intervening cells
liai e conipiessed tlit, sinusoids here, f o r the parenchyma has become swollen as
a r e s d t of the ndniinistratioii of a minimal dose of carbon tetrachloride 24 hours
beforc tlie r a t was sacrificed. Eiicircliiig the central vein are portions of two
tc~riniiial1 enules of tlie portal veiu. r o t e the manner i n which they arborize with
the surrounding siniisoids. X 200.
2 Liver. E:upeiinieiital liistoiy of r a t and details of technical preparation a s
for figure 1. In tlie upper left is a central vein. A portion of a terminal poital
venulc occupies the eentrtil part of the surrounding sinusoids. At the left and
I ~ l o n:ire brandies of a large coiiductiiig portal vein. The slice has been cut in
sucli a niaiiiicr that the non-injected zone surrounding the branch on the left is
cleni lp slion 11, indic:iting tlie absence of direct communication between its lumen
:uid the nd,j;icent siiiusoids. To reacli the latter, iiik (and presumably blood)
travei sed tlic sinusoids intervening hetweeii them and the terminal portal venule
i i i tlic center. Tlie iiijectcd iiih lias been lost from the lumen of the conducting
branch of tlie portal vein in the lower portion of the field. This has revealed that
the sinusoids run over it, but had tlic ink remained tn sttu, the superimposed
sinusoids would ha\e appeared to coinuiunicate with it. X 250.
3 Liver. Rappaport’s structural units have been dissxted by the formation
of fibrous trabeculae in this preparation from a r a t in the early stages of dietary
cirrhosis due to choline deficiency. A large conducting branch of the portal vein
occupies the center of the field. From it radiate a number of terminal portal
renules, each surrounded by a piriform mass of sinusoids supplying parenchyma
which is intact except at the periphery of each acinus. Branches of the central
vein are situated a t the periphery of the field and are linked by a n annular ring
of fibrous tissue from which other trabeculae extend along paths midway between
the terminal venules toward the conducting branch of the portal vein in the center
of the field. I n two places the fibrous tissue has extended to the adventitial sheath
of the conducting vessel. Fibrous trabeculae appear relatively free of injected
ink. Technical preparation as for figure 1. X 100.
4 Liver. This preparation (technic as f o r fig. 1) from a more advanced case
of dietary cirrhosis (choline-deficient r a t ) than that illustrated in figure 3, slioms
two large conducting branches of the portal vein running from left t o right and
terminating just beneath the capsule. Between them (center and extreme bottom
of field) run portions of their corresponding radicles of the central vein. The
latter are completely surrounded by fibrous tissue. The conducting branches of
the portal vein give off numerous terminal venules t o the centers of Rappaport’s
structural units, which have undergone varying degrees of peripheral destruction
and fibrotic replacement. Note that where appreciable amounts of the units hare
become replaced by fibrous tissue, the latter surrounds the corresponding segments of the large portal vein. There is little evidence of fibrosis, however,
immediately adjacent to the terminal venules, indicating that the lesiou is :I
functioiially nonportal type, despite fibrosis around the large portal reins. Further
dcscription given in text. X 100.
9 1.
,5 Liver. This preparation (teclinic as f o r fig. 1) is from an adTaiieed stage
of dietary eirrl’osis in a clioline-deficient rat. Erery large venous trunk, v h e t l m
it be portal or hepatic, is surrounded by fibrous tissue. The only vessels not
embedded in scar tissue are the terminal portal renules wliich are siirrounded by
clumps of sinusoids supplying hyperplastic units of Rappaport. A s the areas
around terminal portal renulcs a r c tlic only lionfibrotic regions left in this lirer,
the lesions may still be regarded a s essentially nonportal ( i n a fiinctioual sense)
despite tlie piesenci: of scar tissue around large portal veins.
6 L i I t r . Paraffin section stained with 1iem:itoxylin and eosin, from a case of
alcoholic cirrhosis in man. The wide h n d of fihlous tissue (left) surrounds
re1:rtively large branches of tlie portal veiu a n d hepatic artery. I n the inidst of
the nodalc of pareiicliymn (estrenie right) is :I t:iiigential section of a terniiii:rl
portal rein and its accom1):iiiyiiig bile ductule. This is free of fibrosis, indicating
tliat, :is i n clioliiie-tieficieiit, cirrhotic rats, t h o lesions arc of a functionally 11011portal nature. X 120.
‘iLiver. Higher magiiification of the terminal portal venule and bile ductule
shown a t the estrenie riglit of figure 6. Note absence of fibrosis. Further deserip
tion in test. X 600.
Без категории
Размер файла
1 561 Кб
unit, anatomy, experimentov, structure, terms, trabecular, dietary, stage, latex, pathogenesis, rappaport, cirrhosis
Пожаловаться на содержимое документа