Introduction

Liver cirrhosis is ultimately caused by diffuse and persistent injury to the liver. Although liver cirrhosis is pathologically defined briefly by regenerative nodules surrounded by diffuse fibrosis, the nodules show a variety of morphological features according to the cause and length of the diseased period. Instead of the morphological variety, it is indicated that there is essentially a common pathway to liver cirrhosis. The pathway is nothing but the process of breakup and rebuilding, namely restructuring, of the liver lobules [1, 9, 10, 16, 17, 22]. However, the precise formal pathogenesis of liver cirrhosis has remained unsolved.

Hence, we tried to clarify the three-dimensional alteration of the liver lobule with special reference to its angioarchitecture in chronic viral hepatitis in order to determine the formal pathogenesis regarding the transformation from chronic hepatitis to liver cirrhosis.

Materials and methods

A liver tissue with chronic viral hepatitis (C type) accompanied with liver cell carcinoma was used. The liver tissue of a 46-year-old Japanese male was resected surgically to remove the tumor. Tissue blocks were taken from the portion sufficiently apart from the tumor. A normal liver of an autopsied case of a 39-year-old Japanese male with cerebral bleeding was used as control. Tissue blocks were embedded in paraffin in the usual manner. Serial sections were cut 3.5 μm in thickness and stained with modified Masson's trichrome. The fundamental structures were outlined serially by free hand on sheets of tracing paper. The drawing on the sheets was facilitated by the use of a projection apparatus attached to a microscope. Completion of reconstruction was made step-by-step by overlapping the outlines.

Although we focused mainly on the portal system in this study, the arterial and biliary systems were also depicted in the drawings in as much detail as possible.

Results

A normal liver

The angioarchitectural reconstruction of the parenchymal portion of the normal liver is shown in Fig. 1 and Fig. 2. The former was 520 μm and the latter 120 μm in thickness. Hereafter, we use the terminology and conception defined by Matsumoto et al. [14, 15] with reference to the angioarchitecture of a normal liver. The portal system is divided into the conducting and parenchymal portions. The main function of the former is the transportation of blood, while the latter provides the angioarchitectural framework of the liver lobule in addition to blood transportation. The portal veins of the parenchymal portion are distinguished by three steps; the first, second and third step branches (Fig. 1). The first step branches arise directly from the terminal conducting portion. The second step branches are given off mainly nearly at a right angle from the first step ones and correspond to the portal veins in the portal tracts located in the periphery of the hexagonal classic lobule. These first and second steps are regularly distributed. The third branches show two main branching modes. One mode is that the third step branches occur nearly at a right angle from the second step ones and run with little connective tissue sheath along the boundary zone of adjacent lobules like the veins of a leaf. The other is a fork-like branching mode at the terminal of the parenchymal portion (Fig. 2). A group of adjoining third step branches formed the basic framework of a sickle-shaped sinusoidal bed (sickle zone) (Fig. 1). The sickle zone is distinguished by a high potential area of blood flow. Several sickle zones cover almost the entire surface of the classic lobule. The portion corresponding to the edge of the sickle zones forms a surface-like inflow front from which blood pours into the sinusoids. It is worth pointing out that the surface-like inflow front shows a systemic continuity throughout the liver (see Matsumoto et al. [14, 15]).

Fig. 1.
figure 1

Histological reconstruction of the normal liver at moderate magnification. The figure covers the parenchymal portion of the portal system composed of three step branches. Note the alternating arrangement of portal and hepatic veins and regular manner of ramification in the portal system. The third step branches form a basic structure of the liver lobule. Adjoining third step branches form a sickle zone with high potential blood flow (hatched line area). 1: First step branch; 2: second step branch; 3: third step branch

Full size image
Fig. 2.
figure 2

Histological reconstruction of the normal liver at high magnification. Third step branches are given off nearly at a right angle in the left half of the figure and fork-like branching mode is shown at the terminal of the portal parenchymal portion in the right half. 2: Second step branches; 3: third step branches

Full size image

A three-dimensional holistic view of the angioarchitecture of the liver with chronic hepatitis by histological reconstruction under a low magnification

The light microscopic features of the case showed enlarged portal tracts with inflammatory cell infiltration and fibrosis, occasional fibrous bridging between the portal tracts or between the portal tract and central vein and scattered focal necrosis and interface hepatitis (Fig. 3). Although the bridging fibrosis suggested the local disorganization of the lobules, the basic structure of the lobules seemed to have a tendency to be maintained on the whole throughout the liver. The complexity of the vasculature was indicated by the increasing number of cut surfaces of blood vessels in the portal tracts.

Fig. 3.
figure 3

Photomicrograph at low magnification. Portal areas are enlarged with lymphocyte infiltration and fibrosis. Fibrous bridging occasionally develops between portal-to-portal tract or hepatic vein. However, basic structures of the liver seem to be preserved in general

Full size image

The reconstruction was 480 μm in thickness and covered the range from the terminal trees of the conducting portion mainly to the second step branches of the portal vein and partly to the third steps (Fig. 4). Fibrous adhesions of the portal to portal tracts or to hepatic veins occasionally occurred and are indicated by arrows in the figure. In general, the first step branches of the portal vein and the corresponding hepatic veins showed a tendency to remain in the normal spatial relationship in that they were equally distributed in number (compare with Fig. 1). However, conspicuous changes in the angioarchitecture began at the second step, or subsequent branches, which showed a loss of regular distribution because they partly decreased and partly increased in number, were branched at various angles, ran in close proximity to each other and were bent sharply. These features resulted in the loss of a regular pattern and complicated appearance of the angioarchitecture. However, the hepatic vein system tended to maintain a basic constellation except for some adherent lesions to the portal tracts.

Fig. 4.
figure 4

Histological reconstruction of chronic hepatitis at low magnification. Spatial arrangement of the distal conducting portion and the first step branches of the parenchymal portion of the portal system are preserved as a rule. Irregular manner of ramification begins roughly at the second step or subsequent branches of the portal vein. Fibrous bridging lesions are found between the portal to portal tracts (arrows) or hepatic veins (double arrows). Area with an increase in the number of branches is indicated by double arrowheads. A single arrowhead shows an abnormal bend of the portal vein. CP conducting portion, 1: first step branch; 2: second step branch; 3: third step branch

Full size image

These findings of the altered constellation of blood vessels appear to indicate the early phase of the loss of the basic framework of the lobular angioarchitecture, which was difficult to demonstrate in a histological plane section.

The reconstruction of the portal area corresponding to the parenchymal portion with relatively less damage

Common features in this case were the enlarged portal tracts accompanied by inflammation and fibrosis without fibrous bridging (Fig. 5). The reconstruction of this area was 280 μm in thickness and covered the terminal branch of the conducting portion and all the step branches of the parenchymal portion (Fig. 6). The second step branch b was a distal branch of a. Although the angioarchitecture seemed to remain normal as a rule, a precise examination revealed that branching of b portal vein was severely distorted. It gave off many more branches, corresponding to third step ones, than are normal (arrowheads in Fig. 6; compare with Fig. 2). This suggested that these branches contained not only preexisting branches, but also newly formed ones. Additionally, they were arranged in a disorderly fashion. Nearly half of them could not be traced because they were stenosed and buried in the connective tissue (arrows in Fig. 6). The remainder ran into the sinusoid to perfuse the parenchyma (asterisks in Fig. 6).

Fig. 5.
figure 5

Photomicrograph of histology of the reconstructed area demonstrated in Fig. 6. The portal area is inflamed and enlarged by fibrosis. Note the complexity of vasculature. Original magnification ×13

Full size image
Fig. 6.
figure 6

Reconstruction of chronic hepatitis at moderate magnification. Angioarchitecture seems to be maintained as a rule. The number of third step corresponding branches increases and they run randomly (a part is indicated by arrowheads). While some of the third step branches are buried and disappear (thick arrows), the remaining ones reached the parenchyma to perfuse the cells (asterisks). Compare with Fig. 1. Branch b is a distal branch of a. CP: conducting portion; 1: first step; 2: second step; 3: third step

Full size image

The reconstructions of the angioarchitecture in the fibrous bridging lesions under a high magnification

The portal-central (p-c) fibrous bridging lesions in the parenchymal portion

A representative light microscopic picture of the lesions is shown in Fig. 7. The reconstruction of the angioarchitecture was 420 μm in thickness and mainly included the parenchymal portion of the portal vein (Fig. 8). The enlarged portal tracts b, c and d were connected with hepatic veins equivalent to the first step and its distal branches by irregular fibrous bands. Initially, the blood vessel arrangement appeared to be apparently unlike that of a normal liver in each area. In area b, it was not easy to identify the order of the portal branches. The portal veins, which were speculated to be the second step and its distal branches, ran in the fibrosed area. They gave off only a small number of branches on their course, resulting in a decrease of branches. Some branches were stenosed and buried in fibrosed tissue (arrows in Fig. 8), while the remainder reached the inside of the lobule (asterisks in Fig. 8). In area c, it was evident that the second step and its distal branches ran relatively randomly. Two portal veins were found to have run rather a long distance without effectively branching to the parenchyma and they finally disappeared with stenosis in the connective tissue (arrowheads in Fig. 8). Although area a seems to have a tendency to maintain the normal ramification pattern of portal veins, the first and second step branches became thicker and took tortuous courses. In more detail, while the branches equivalent to the third step branches were frequently found in parts, the portal vein ran simply without ramification in other parts as seen in area c (arrowhead in Fig. 8). The hepatic veins showed a distortion of their course and an increase in thickness; however, their terminal trees expanded in the lobule to drain sinusoidal flow. Thus, the drainage system was less damaged than the portal system.

Fig. 7.
figure 7

Photomicrograph of the reconstructed area demonstrated in Fig. 8. Portal areas are inflamed and enlarged by fibrosis. Portal areas are connected by fibrosis with the hepatic vein (arrow). Paucity of vessels is conspicuous in the fibrosed areas. Original magnification ×12

Full size image
Fig. 8.
figure 8

Reconstruction of the lesion with p-c bridging fibrosis. Portal areas b, c and d are connected with the hepatic vein. Preservation of the portal tree varies from place to place. Portal area b connected with the hepatic vein shows a distal portal branch giving off only a small number of branches, one of which run into the parenchyma (asterisk) and the other of which disappears in the fibrosed area (arrow). Other asterisks and arrows indicate the same things in other areas. In portal areas a and c, a few portal branches are found to run for a relatively long distance without giving off effective branches to the parenchyma (arrowheads). Contrary to the portal system the hepatic vein extends out abundantly in the parenchyma. 1: first step; 2: second step; 3: third step

Full size image

The fibrous bridging lesion of the portal area to the hepatic vein corresponding to the conducting portion

A photomicrograph of the lesion and figure of its reconstruction are shown in Fig. 9 and Fig. 10. The latter was 670 μm thick. In the reconstructed area, roughly four fibrosed lesions (a, b, c and d) were observed to adhere to a thick hepatic vein corresponding to the conducting portion. Areas a and b were rich in vasculature, while the remainder showed a paucity of blood vessels. Areas a, b and d were identified as altered portal tracts because of the presence of portal veins that appeared most likely to be second step and their distal branches. Area c was also speculated to be fibrosed portal tract. The portal veins, arteries and hepatic veins were tangled with each other like ivy in area a. The portal veins showed abnormalities such as parallel running of two portal veins of the same order, slight tortuosity and paucity of branching. A few branches reached the edge of the parenchyma to nourish the cells (asterisk in Fig. 10). In the same area, hepatic veins were also found twisting their way through the other vessels. A normal angioarchitecture was not found. The intermingling of various vessels in the fibrous area resembled liver cirrhosis. These features indicated almost complete loss of lobular architecture. In area b, portal veins showed a ramification pattern similar to a normal ramification pattern to some extent. Nearly half of the branches equivalent to the third step ran into the parenchymal region (asterisks in Fig. 10). In both portal areas a and b, the development of arteries was conspicuous. The arteries sometimes ran over a long distance. However it was difficult to find branches that terminated clearly in the parenchyma to nourish the liver cells. Areas c and d with prominent fibrosis were speculated to be brought about by severe damage to the parenchyma containing the small portal tracts and replacement by fibrosis, although portal veins were seldom seen. Although the hepatic vein of the conducting portion showed slight irregularity of caliber and distortion of running course by the fibrous adhesion in part, the distal branches spread abundantly in the lobule to drain blood.

Fig. 9.
figure 9

Photomicrograph of the reconstructed area demonstrated in Fig. 10. Portal tracts and conducting portion of the hepatic vein (arrows) are fused together. Portal areas are inflamed and poor in vasculature. Original magnification ×14

Full size image
Fig. 10.
figure 10

Reconstruction of the severely damaged lesion with p-c bridging fibrosis. Four fibrosing areas (a, b, c and d) are observed to be adherent to the hepatic vein corresponding to the conducting portion. Areas a and b are rich in vascularity and show intermingling of portal veins, arteries and hepatic veins. Although complete loss of lobular angioarchitecture is indicated in areas a and b, some branches reach into parenchyma (asterisks). Areas c and d are poor in vasculature. Proliferation of bile ductules is seen

Full size image

In the range of this histological reconstruction, the angioarchitecture was severely altered in general in comparison with the previous reconstructed area. The normal lobular angioarchitecture almost disappeared, especially in the fibrosed area around the large vein with a few exceptions.

The portal-portal (p-p) fibrous bridging lesions

As indicated by the histological reconstruction mentioned above, it is usual for diseased portal areas to be connected to each other by fibrous tissue with a paucity of vasculature. Paucity of vasculature or absence of thick portal veins indicates that the fibrous bridging area involved mainly distal portal tracts which were accompanied by the disappearance or marked decrease of the portal branches and parenchyma caused by necroinflammatory reaction. In comparison to p-c bridging lesions, the severity of the loss of parenchyma was thought to be less in p-p lesions.

Discussion

The aim of this report was to clarify the exact alteration of the lobular architecture in chronic hepatitis as much as possible to determine the formal pathogenesis of cirrhotic liver. The lobule is an essential concept concerning structure and indicates unitary parenchymal mass. Until now, three representative concepts of the liver lobule have been proposed [7]. They are "classic lobule" by Kiernan [8], the portal lobule by Mall [12] and the liver acinus by Rappaport [24]. With regard to this, Matsumoto's concept has already been described as a subitem of a normal structure.

We adopted the concept of the hexogonal classic lobule in this study because the concept is more reasonable for explaining the pathological changes, for example, the circulatory disturbances of the liver.

The angioarchitecture of the cirrhotic liver has been investigated by many researchers using the injection technique since the early twentieth century. Kretz [10], McIndoe [17], Popper et al. [23], Mann et al. [13], Nakamura et al. [20] and others showed the distortion of the angioarchitecture and anastomoses between portal and hepatic veins in cirrhotic livers with blood vessels treated with the injection cast technique. Mitra [18] reported a similar distorted angioarchitecture, however, failed to find anastomoses. In addition, Moschcowitz [19] also discussed the process of fibrosis in Lannec liver cirrhosis.

On the basis of pathological findings, the structural alterations of liver cirrhosis were variously described such as "Umbau", rebuilding, by Kretz [9], distortion of the lobule by Moschcowitz [19], altered reconstruction of the lobular pattern by Popper [22], abnormal reconstruction of the preexisting lobular architecture by Popper [21] and disorganization of the lobular architecture by Anthoney et al. [1]. Needless to say, these expressions meant the development of the newly formed parenchymal mass (regenerative nodules) was replacing the normal lobules. However, the precise process of the lobular architecture, rebuilding, has remained as yet unsolved.

We are now facing the problem that Popper et al. [22] stated in 1955—that the most acceptable description of cirrhosis is that of an altered reconstruction of the lobular pattern. Some Japanese literatures with English abstracts have reported the altered angioarchitecture of cirrhotic livers by histological reconstruction methods using serial tissue sections [5, 6, 11]. The results of these studies were briefly summarized as follows. Cirrhotic liver showed the complete loss of the normal lobular angioarchitecture due to severe damage to portal trees. This meant the loss of continuity of the surface-like inflow-front throughout the liver. Therefore, new blood supply and a venous draining system developed in regenerative nodules. Namely, inflow sources changed to point-like ones from which blood irrigated regenerative nodules. Draining veins corresponding to point-like inflow sources were distributed in the same nodules. As a result, regenerative nodules demonstrated conspicuous complexity concerning blood supply and venous draining system. These findings conflict with the view on the pathogenesis of cirrhosis by Rapport et al.—that nodules are remnants of acini of various orders that had their zonal damage replaced by scar tissue [25].

We will primarily discuss the distortion of the lobular architecture based on altered angioarchitecture of our case with chronic hepatitis, taking into account the problem of rebuilding of lobules.

The initial damage of preexisting portal veins themselves on microscopic examination was stenosis and disappearance with the consequence of complexity of vasculature partly due to new formation of distal portal veins. The damage was considered to be caused by the involvement of portal veins in inflammation. In addition, the persistence of inflammation and fibrosis following necroinflammatory lesion [2 ,3, 4] in parenchyma containing portal tracts was deemed to induce subsequent new formation of portal veins, abnormal running course such as sharp bends and close apposition of portal veins of the same order. These changes resulted in spatial disarrangement of portal veins and hepatic veins. The severity of these changes is presumed to depend on the duration, intensity and deterioration of inflammation.

As mentioned above, various types of damage to the angioarchitecture were revealed by histological reconstruction. There is no doubt that the alteration of blood vessels, especially portal veins of the parenchymal portion, more or less affects the angioarchitectural skeleton of normal lobules. This results in the alteration of the lobular structure. Although the severity varied from place to place, the distortion began mostly from the second step or subsequent branches. The mild distortion is shown in Fig. 6, while the severely damaged areas are demonstrated in Fig. 8 and Fig. 10. In the former, the possibility of the lobular angioarchitectural framework partially remaining was indicated. However, in the latter, especially in Fig. 10, conspicuous spatial derangement of blood vessels apparently suggests its complete loss. It is worth pointing out that the coexistence of portal and hepatic veins in the same fibrous septa usually seen in liver cirrhosis had already developed in the advanced lesions of chronic hepatitis. In other words, this indicates an early histological sign of cirrhosis. In this manner, the lobular architecture is considered to be altered step-by-step with the destruction of angioarchitecture and to be completely lost ultimately throughout the liver as the disease progresses. This process of destruction of the lobular angioarchitecture is conceived to be common pathway to liver cirrhosis regardless of its cause.

How was the blood supply to parenchymal cells achieved in the areas with the distortion of angioarchitecture? It is well known that the liver depends mainly on portal blood. It was also pointed out by Mitra that the parenchymal cells of cirrhotic liver were supported by portal blood [18]. Our study suggested that the same could be applied to chronic hepatitis. Namely, it was conceivable that only the remaining third step branches or newly formed ones supplied portal blood to the parenchymal area. Then, inflow source has inevitably changed from a surface-like to point-like inflow source in part in the damaged areas. The important thing is that these remaining portal veins containing point-like inflow sources provide a new nutritional basis for liver cells to regenerate. This is also one of the initial conditions of the development of liver cirrhosis. Although the marked development of arteries was observed, especially in the severely damaged lesions, they were considered to make far less of a contribution to maintaining the parenchyma in comparison to portal vessels—contrary to the opinions by Kretz [10] and McIndoe [17].

Although we failed to find anastomosis between portal and hepatic veins in this case, there is a possibility of its presence in the bridging fibrosed areas.

Why did the liver of this case still remain in the state of chronic hepatitis, but not liver cirrhosis, despite diffuse damage to the microcirculation? The principal reason is considered to be that damage to the parenchyma is uneven, and mild to moderate damage dominates throughout the liver, although severely damaged parenchyma is sometimes found. It is also possible that the areas with a tendency to retain the basic lobular angioarchitecture prevailed, although not completely, in this case. This point of chronic hepatitis is a great difference from a cirrhotic liver, which shows a complete replacement of the liver parenchyma by regenerative nodules and almost complete loss of a normal lobular angioarchitecture.

We have observed the changes in the lobular angioarchitecture in chronic hepatitis compared with a normal liver. It is evident that chronic hepatitis is situated in the process toward cirrhosis and already contains early signs of liver cirrhosis. It is relatively easy to presume that chronic hepatitis finally progresses to liver cirrhosis as a result of the persistence of inflammation, fibrosis and regeneration over a long period. However, we only clarified a part of the wide spectrum of the process. Further investigation is needed to elucidate the formal pathogenesis of liver cirrhosis.