Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print March 13, 2007, 10.1148/radiol.2432060341
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2432060341v1
243/2/386    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Liu, Y.
Right arrow Articles by Matsui, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Liu, Y.
Right arrow Articles by Matsui, O.
(Radiology 2007;243:386-395.)
© RSNA, 2007

Experimental Studies

Changes of Intratumoral Microvessels and Blood Perfusion during Establishment of Hepatic Metastases in Mice1

Yi Liu, MD2 and Osamu Matsui, MD

1 From the Department of Radiology, Kanazawa University, Graduate School of Medical Science, 13-1 Takara-machi, Kanazawa 920-8641, Japan. Received February 22, 2006; revision requested April 25; revision received May 25; accepted June 20; final version accepted August 14. Supported in part by the Ministry of Education, Science, Sports and Culture, Grant-in-Aid for Scientific Research (C), 16591192, 2004, and by a Grant-in-Aid for Cancer Research (15) from the Ministry of Health, Labour and Welfare. Address correspondence to O.M. (e-mail: matsuio{at}med.kanazawa-u.ac.jp).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To prospectively evaluate the stepwise changes that occur in intratumoral microvessels and microcirculation during the establishment of murine colonic hepatic metastases by using in vivo fluorescent microscopy and to compare the changes with tumor angiogenesis evaluated with an immunohistochemical study.

Materials and Methods: This study was approved by the institutional animal care and use committee. Twenty-five mice with hepatic metastases created with injection of murine colonic adenocarcinoma (colon 26) tumor cells into the spleen were examined with in vivo microscopy and immunohistochemical study for CD34, intracellular adhesion molecule (ICAM-1), and alpha smooth muscle actin ({alpha}-SMA). The tumor size, microcirculation in tumors, intratumoral microvessel density (MVD), afferent MVD, and CD34-positive MVD were evaluated. The data among the tumors that showed different hemodynamic or immunohistochemical patterns were compared with the Kruskal-Wallis test and the Student t test.

Results: Four stepwise patterns were observed according to the changes in morphology, hemodynamics, and immunohistochemical characteristics of intratumoral microvessels during the establishment of hepatic metastases, as follows: metastases without definite intratumoral blood perfusion or any intratumoral microvessels (mean diameter, approximately 180 µm), metastases with portal perfusion and intratumoral ICAM-1–positive residual hepatic sinusoids (mean diameter, approximately 290 µm), metastases with mixed portal and arterial perfusion and increased CD34-positive microvessels and {alpha}-SMA–positive arterioles (mean diameter, approximately 520 µm), and metastases with exclusively arterial perfusion and increased CD34-positive microvessels and {alpha}-SMA–positive arterioles (mean diameter, >2000 µm). The differences among the mean sizes of the tumors that showed these four patterns were statistically significant (P < .01).

Conclusion: Stepwise changes of intratumoral microcirculation were revealed from direct diffusion, to portal perfusion, to mixed portal and arterial perfusion, and finally to arterial perfusion in accordance with stepwise tumor neovascularization during the growth of murine colonic hepatic metastases.

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
The liver is a common site of metastases from various primary tumors, and such hepatic metastases are an important prognostic determinant (1). In the past decade, several methods, which include vascular cast, quantitative autoradiography, dynamic computed tomography (CT), Doppler ultrasonography, and in vivo microscopy (2–10), have been used to define the microvascular architecture and blood perfusion of hepatic metastases at various stages. Some researchers have demonstrated the value of in vivo microscopy when it is applied to the study of the microcirculation of hepatic metastases (7–10)—that is, the real-time observation of blood flow and tumor perfusion at an early stage of these metastases without an effect on natural hemodynamics. With this method, the morphologic and hemodynamic features of hepatic metastases have been evaluated, but conflicting results have also been reported. In our opinion, these discrepancies were mainly caused by the lack of serial in vivo observation of the morphologic and hemodynamic changes of intratumoral microvessels during the establishment of liver metastases.

Thus, the purpose of our study was to prospectively evaluate the stepwise changes that occur in intratumoral microvessels and microcirculation during the establishment of murine colonic hepatic metastases by using in vivo fluorescent microscopy and to compare the changes with tumor angiogenesis evaluated with an immunohistochemical study.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Animals
Twenty-five female BALB/c mice (Sankyo Laboratory, Toyama, Japan), aged 8–12 weeks and weighing 20–30 g each, were included in this study. This study was approved by the animal care and use committee of our university and was performed in accordance with its guidelines.

Tumor Model
A murine colonic adenocarcinoma cell line (colon 26), which is commonly used in various studies of murine hepatic metastases (11), was maintained in Dulbecco's modified Eagle's medium (Nissui Pharmaceutical, Tokyo, Japan) with 10% heat-inactivated fetal bovine serum at 37°C in a humidified atmosphere with 5% CO2. Subconfluent cells were collected and resuspended in Hanks balanced salt solution (Sigma Chemical, St Louis, Mo) at a cell density of 5 x 105 cells per milliliter. One individual (Y.L.) injected 100 µL of cell suspension into the spleen in 25 mice. The animals were anesthetized with light ether.

In Vivo Study
In vivo microscopy by using a fluorescent microscope (Olympus BX41; Olympus Optical, Tokyo, Japan) was performed on exteriorized livers 8–17 days after intrasplenic inoculation of colon cancer cells (Table 1). The morphologic and hemodynamic changes of intratumoral microvessels were observed according to reported techniques (12). To evaluate the intratumoral portal supply, 1.68 x 107 fluorescent microspheres (Fluoresbrite plain YG; Polysciences, Warrington, Pa) with a diameter of 3 µm that were suspended in 0.1 mL saline were injected into a branch of the portal vein in eight mice during in vivo microscopy by an individual (Y.L.). To analyze the intratumoral arterial supply, the fluorescent microspheres were injected into the tail vein (Y.L.) in six other mice. This site of injection was chosen because the selective injection of microspheres into the hepatic artery by means of retrograde cannulation into that artery from the gastroduodenal artery, to keep its natural blood flow intact, was technically impossible. When these microspheres that were injected into the tail vein appeared in the tumor earlier than or almost simultaneously with those in the surrounding hepatic sinusoid, and then drained into the surrounding hepatic sinusoids, the supply was considered an arterial blood supply. The other mice (n = 11) without injection of fluorescent microspheres were observed by the same individual who performed the injection. In each mouse, several tumors were observed, but only one that was selected randomly could be observed during injection of fluorescent microspheres. All mice underwent a single session of in vivo microscopy and then were sacrificed with an overdose of anesthetic. Microscopic images were recorded at a rate of 30 images per second with a digital video camera (DXC-108; Sony, Tokyo, Japan) and were transferred to a digital video system (NV-DM1; Panasonic, Tokyo, Japan) for off-line analysis.


View this table:
[in this window]
[in a new window]

  		Table 1. Schedule of in Vivo Microscopy in Mice

 
Histologic and Immunohistochemical Staining
Twenty-five livers were removed and fixed in 10% formalin after in vivo microscopy; they were then embedded in paraffin, and 5-µm serial sections were cut for hematoxylin-eosin and immunohistochemical staining. For immunohistochemical staining performed by the same individual who performed the injection as previously mentioned, the sections near the center of each metastasis were chosen. After deparaffinization, the sections were incubated in methanol with 0.3% H2O2 for 30 minutes to remove endogenous peroxidase. Subsequently, nonspecific protein binding was blocked with 10% serum obtained from the same species used to produce the secondary antibody. Thereafter, the sections were incubated overnight at 4°C with one of the following primary antibodies: (a) rat antimouse CD34 (HyCult Biotechnology, Uden, Netherlands), a specific marker for vascular endothelium and not expressed on the normal hepatic sinusoidal epithelium, diluted at a ratio of 1:50 (13,14) for evaluation of vascularization in the tumor and (b) rat antimouse intracellular adhesion molecule (ICAM-1) (BioLegend, San Diego, Calif), usually expressed on the hepatic sinusoidal epithelium but not on the newly formed vessels, diluted at a ratio of 1:50 (15,16) for analysis of the existence of hepatic sinusoids in the tumor and surrounding it. After the sections were washed three times with phosphate-buffered saline, they were incubated for 1 hour at room temperature with biotinylated rabbit antirat IgG (Vector Laboratories, Burlingame, Calif) diluted at a ratio of 1:100, after which the immune complexes were visualized by using a kit (Elite ABC; Vector Laboratories). For other sections, the rabbit antimouse alpha smooth muscle actin ({alpha}-SMA) (Spring Bioscience, Fremont, Calif), diluted at a ratio of 1:100, was incubated for 15 minutes at room temperature to evaluate intratumoral arterial neovascularization. Then, the sections were treated with an immunohistochemical staining system (EnVision [+] system with diaminobenzidine; Dako/Cytomation, Carpinteria, Calif) according to the manufacturer's instructions to visualize the vascular smooth muscle.

Image Analysis
In vivo images of 82 tumors were downloaded onto a computer (NetVista A30; IBM, New York, NY) equipped with software (Windows 2000; Microsoft, Redmond, Wash) from videotapes by using a real-time digital video image capture card (EZDV2; Canopus, Kobe, Japan) and were analyzed with imaging software (Osiris; Digital Imaging Unit, University Hospital of Geneva, Geneva, Switzerland) and the software for quantitative analysis of angiogenesis network (Angiogenesis Image Analyzer; Kurabo Industries, Osaka, Japan). On in vivo images, the following data were collected by two individuals (Y.L. and O.M.): (a) tumor size, which was defined as the maximal diameter of the tumor on the liver surface; (b) intratumoral microvessel density (MVD), which was defined as a ratio of the areas between the intratumoral microvessels and tumor on the liver surface and was calculated as the total area of observed intratumoral microvessels in one tumor divided by the extent of this tumor on the liver surface; and (c) afferent MVD, which was evaluated to compare the amount of direct portal blood perfusion to metastases, defined as the number of afferent microvessels per millimeter on the circumference of the tumor, and was calculated as the number of afferent microvessels in one tumor divided by the length of its circumference on the liver surface.

On immunohistochemical sections, data were collected by two individuals (Y.L. and O.M.). The steps involved in the collection were as follows: The maximal diameter of hepatic metastases was employed to determine the tumor size by using the imaging software named previously. The intratumoral CD34-positive MVD, which was defined as the number of CD34-positive intratumoral microvessels, was estimated at three high-power fields on the region of tumor with the most microvessels. In addition, the intratumoral {alpha}-SMA–positive arterioles were defined as the intratumoral vessels that were positive for anti–{alpha}-SMA staining without accompanying portal veins or bile duct. The ratio between the thickness of the medial layer that shows {alpha}-SMA–positive staining and the minimal external diameter of this vessel should be more than 1/10 to exclude the positive staining of venous components or perisinusoidal stromal cells (17,18).

Statistical Analysis
Data were initially assessed for normality with use of normal probability plots and were presented as the mean ± standard deviation. On the basis of these results, nonparametric procedures were used to compare the tumor size of metastases showing different hemodynamic or immunohistochemical patterns. Given the independent and unpaired nature of these data, the Kruskal-Wallis test was initially used to test overall equality of medians in each data group. When a statistically significant difference was observed, single posttest comparisons of independent samples were performed by using the Mann-Whitney test. An overall difference with P = .05 was considered significant. A two-tailed unpaired Student t test was used to determine whether significant differences existed in the MVD, afferent MVD, and CD34-positive MVD among each group, which showed different hemodynamic or immunohistochemical patterns. The difference was considered to be statistically significant if the P value was less than .05. All statistical analyses were performed with software (SPSS, version 10.0, 1999; SPSS, Chicago, Ill).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
All mice survived the tumor cell inoculation and the in vivo microscopic observations. In vivo microscopic observation of the morphologic and hemodynamic changes of intratumoral microvessels was performed in 82 variously sized hepatic metastases; 126 lesions were used to evaluate the changes of tumor vascularization at the immunohistochemical study.

Evaluation of Morphologic and Hemodynamic Changes of Intratumoral Microvessels with in Vivo Fluorescent Microscopy
Four stepwise patterns were observed in regard to the morphologic characteristics of intratumoral microvessels and microcirculation of the tumor and the surrounding hepatic sinusoids at in vivo microscopy (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 2. Size of Hepatic Metastases according to in Vivo Microscopic and Immunohistochemical Patterns

 
Pattern I: hepatic metastases without definite intratumoral microvessels.—At the very early stage, hepatic metastases were limited to clusters of tumor cells. Neither intratumoral microvessels nor blood perfusion to the hepatic metastases was observed (Fig 1). Because there were no definite intratumoral microvessels, a study with injection of fluorescent microspheres was not performed in these tumors. The mean size of the tumors that showed this pattern was 179.5 µm ± 25.4.


Figure 1
View larger version (77K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Hepatic metastases without definite intratumoral microvessels. Top: Schematic diagram. Bottom: In vivo fluorescent microscopic image. At the very early stage, hepatic metastases were limited to clusters of tumor cells in the liver with no intratumoral microvessels, whereas the hepatic sinusoids adjacent to the tumor (*) showed slight dilatation and deformity. Arrows indicate blood flow direction. Bar = 100 µm, S = sinusoids, T = tumor, THV = terminal hepatic venules, TPV = terminal portal venules.

 
Pattern II: hepatic metastases with hepatic sinusoidlike intratumoral microvessels that show direct portal perfusion.—With growth of metastases, microvessels began to appear in the tumor. These intratumoral microvessels were similar to the surrounding hepatic sinusoids in shape, and they anastomosed with the sinusoids. Fluorescent microspheres, which were injected into the portal vein (two mice) or tail vein (two mice), reached the hepatic sinusoidlike intratumoral microvessels through the surrounding hepatic sinusoids, passed through them, and then drained into the surrounding hepatic sinusoids again (Fig 2). These phenomena indicate that at this stage hepatic metastases derive their blood supply directly from the portal vein. The mean size of the tumors that showed this pattern was 287.3 µm ± 48.7.


Figure 2
View larger version (72K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Hepatic metastases with hepatic sinusoidlike intratumoral microvessels (V) that show direct portal perfusion to them. Top: Schematic diagram. Bottom: In vivo fluorescent microscopic image. At this stage, intratumoral microvessels began to appear in the metastases. These intratumoral microvessels were similar to the surrounding sinusoids in shape, and formed anastomoses with them, with the blood flow in them coming from the neighboring sinusoids (portal vein) and draining to these neighboring sinusoids again. Arrows indicate blood flow direction. Bar = 100 µm. Remaining keys are the same as on Figure 1.

 
Pattern III: hepatic metastases with convoluted intratumoral microvessels that show double blood supply.—When the metastases became larger, the intratumoral microvessels increased and they were distributed in a spatially heterogeneous fashion. In addition, they appeared tortuous and developed irregular branching and abrupt diameter changes to form irregularly dilated blood spaces. These convoluted microvessels formed irregular networks in the tumor. Fluorescent microspheres injected into the portal vein (four mice) immediately reached some of the intratumoral microvessels through the surrounding hepatic sinusoids. On the other hand, the fluorescent microspheres injected from the tail vein (two mice) appeared in the tumor and the surrounding hepatic sinusoids almost simultaneously, after which some of the microspheres that flowed into the tumoral parenchyma drained into the surrounding hepatic sinusoids again (Fig 3). These findings mean that, at this stage, the hepatic metastases derived their blood supply from both the hepatic artery and the portal vein. The mean size of the tumors that showed this pattern was 518.6 µm ± 125.5.


Figure 3
View larger version (78K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Hepatic metastases with convoluted intratumoral microvessels that show double blood supply. Top: Schematic diagram. Bottom: In vivo fluorescent microscopic image. At this stage, the intratumoral microvessels increased and were distributed spatially in a heterogeneous fashion. Some irregularly dilated blood spaces (*) appeared. Blood flow from the neighboring sinusoids was observed. Arrows indicate blood flow direction. Bar = 200 µm, DBS = dilated blood space. Remaining keys are the same as on Figures 1 and 2.

 
Pattern IV: hepatic metastases with convoluted intratumoral microvessels that show exclusively arterial blood supply.—With further growth to massive size, although the intratumoral microvessels remained tortuous and showed irregular branching, they became regular, and they were distributed evenly. Almost all the blood flow in the peripheral intratumoral microvessels drained into the neighboring hepatic sinusoids. Fluorescent microspheres injected into the portal vein (two mice) were absent in the tumoral parenchyma regardless of the dense inflow into the surrounding hepatic sinusoids. In contrast, the fluorescent microspheres injected from the tail vein (two mice) appeared almost simultaneously in the intratumoral microvessels and surrounding hepatic sinusoids. Finally, the microspheres that flowed into the tumor drained into the surrounding hepatic sinusoids again (Fig 4). These findings mean that, at this stage, the hepatic metastases derived their blood supply directly from the hepatic artery, and the blood that flowed into the tumor drained into the surrounding hepatic sinusoids through the abundant connections between intratumoral microvessels and the surrounding hepatic sinusoids. All four tumors that showed this pattern were larger than 2000 µm in diameter.


Figure 4
View larger version (81K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Hepatic metastases with convoluted intratumoral microvessels that show exclusively arterial blood supply. Top: Schematic diagram. Bottom: In vivo fluorescent microscopic image. At this stage, the intratumoral microvessels were tortuous and showed irregular branching. Almost all of the blood flow in the peripheral intratumoral microvessels that communicated with neighboring sinusoids drained to the neighboring sinusoids, which was shown by the fact that the fluorescent microspheres were absent in the tumor at the beginning of the injection into the branch of portal vein. Arrows indicate blood flow direction. Bar = 400 µm. Keys are the same as on Figures 1 and 2.

 
The differences among the mean sizes of the tumors that showed the different patterns were statistically significant (P < .01) (Table 2).

The hepatic sinusoids adjacent to the tumors were compressed, were infiltrated, and showed slight deformity at in vivo microscopy (Figs 1–4).

Intratumoral MVD and Afferent MVD
To evaluate the intratumoral MVD and afferent MVD of hepatic metastases, 71 lesions that showed patterns II, III, and IV were used (Table 3). The mean intratumoral MVD of metastases that showed pattern II was 11.9% ± 2.6, which was less than that of metastases that showed patterns III and IV (17.1% ± 3.2 and 15% ± 0.8, respectively; P < .01). On the other hand, the mean afferent MVD of metastases that showed pattern II was 3.46 microvessels per millimeter ± 0.39, which was higher than that of metastases that showed pattern III (3.04 microvessels per millimeter ± 0.58, P < .01), and no definite afferent microvessels were observed in the nodules that showed pattern IV. These findings mean that tumor neovascularization increased but portal supply decreased in accordance with the development of metastases.


View this table:
[in this window]
[in a new window]

  		Table 3. MVD of Hepatic Metastases according to in Vivo Microscopic and Immunohistochemical Patterns

 
Immunohistochemical Study
Immunohistochemical staining with anti-CD34 and ICAM-1 antibodies revealed three stepwise patterns of intratumoral microvessels during the development of hepatic metastases (Table 2).

Pattern 1: hepatic metastases without intratumoral microvessels.—At the very early stage, only clusters of tumor cells were present, and they were distributed mainly at the peripheral portion of liver. Staining for both CD34 and ICAM-1 was negative in the tumor parenchyma (Fig 5). The mean size of the tumors that showed this pattern was 185.9 µm ± 30.1; the mean size was similar to that of metastases that showed pattern I at the in vivo study.


Figure 5
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: Hepatic metastases without intratumoral microvessels. Immunohistochemical staining for CD34 and ICAM-1 (bottom). At the very early stage of hepatic metastases, staining for both CD34 and ICAM-1 was negative in the tumor cell groups. Bar = 100 µm. (Original magnification, x400.)

 
Pattern 2: hepatic metastases with ICAM-1–positive microvessels.—With the growth of metastases, there were ICAM-1–positive microvessels in the tumor, which communicated with neighboring ICAM-1–positive sinusoids. These ICAM-1–positive intratumoral microvessels were considered to be entrapped residual hepatic sinusoids in the tumor. However, a few CD34-positive intratumoral microvessels appeared at this stage (Fig 6). The mean size of the tumors that showed this pattern was 282 µm ± 42; the mean size was similar to that of metastases that showed pattern II at the in vivo study.


Figure 6
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6: Hepatic metastases with ICAM-1–positive microvessels. Immunohistochemical staining for CD34 (top) and ICAM-1 (bottom). At this stage, a few CD34-positive intratumoral microvessels appeared, and for ICAM-1, microvessels began to appear in the metastases. There were ICAM-1–positive microvessels (arrowheads) in the tumor that communicated with neighboring sinusoids. Bar = 100 µm. (Original magnification, x400.)

 
Pattern 3: hepatic metastases with CD34-positive microvessels.—When the metastases became larger, there were many CD34-positive microvessels in the tumor. They were distributed unevenly, and some CD34-positive dilated blood spaces were found. Meanwhile, the ICAM-1–positive microvessels disappeared from the center and were distributed to the edge of the tumor, and they communicated with both the intratumoral microvessels and the neighboring ICAM-1–positive hepatic sinusoids (Fig 7). The mean intratumoral CD34-positive MVD in these metastases was 25.2 microvessels per high-power field ± 4.5, which was much higher than that of metastases that showed pattern 2 (13.1 microvessels per high-power field ± 1.4, P < .01). The mean size of the tumors that showed this pattern was 832.4 µm ± 496.7.


Figure 7
View larger version (183K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7: Hepatic metastases that contained CD34-positive microvessels. Immunohistochemical staining for CD34 (left) and ICAM-1 (right). At this stage, there were many CD34-positive microvessels in the tumor. ICAM-1–positive microvessels were distributed at the edge of the tumor. (Original magnification, x100.) Bottom right inset: Some CD34-positive dilated blood spaces were found. Top right inset: Microvessels communicated with both the intratumoral microvessels and the neighboring sinusoids. Bar = 200 µm. (Original magnification [both insets], x400.)

 
The differences among the mean sizes of the tumors that showed the different patterns were statistically significant (P < .01) (Table 2).

The intratumoral {alpha}-SMA–positive arterioles were irregular, were tortuous, and had various diameters. They were found in 21 metastases, and all of them, including the four massive lesions observed with in vivo microscopy, showed pattern 3 at immunohistochemical study. The mean size of these lesions was 1343 µm ± 634, and the sizes ranged from 497 to 2632 µm (Fig 8).


Figure 8
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8: Immunohistochemical staining shows intratumoral {alpha}-SMA–positive arterioles that were found in large metastases. They appeared irregular, were tortuous, and had various diameters. Bar = 500 µm. (Original magnification, x40.) Top left and top right insets: Small intratumoral {alpha}-SMA–positive arterioles. (Original magnification, x400.)

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
It has been well established that clinically detectable metastases receive a blood supply exclusively from the hepatic artery, with an almost completely absent intratumoral portal blood supply, as revealed with CT during arterial portography and with CT during hepatic arteriography (19,20). On the other hand, there are several reports in which the portal blood supply to clinical (21) or experimental (7) hepatic metastases is described. To resolve this discrepancy, it is important to understand the pathophysiologic characteristics of hepatic metastases.

We evaluated the stepwise morphologic and hemodynamic changes of intratumoral microvessels in vivo, for the first time to our knowledge, in experimental liver metastases created with the injection of colon 26 tumor cells into the spleen in mice. We also classified the changes into four patterns according to the developmental steps from a minute microscopic focus to a macroscopic focus.

The first pattern of hepatic metastases (mean diameter, approximately 180 µm) was without definite intratumoral microvessels and blood perfusion; the metastases were both CD34 and ICAM-1 negative. The metastases were just clusters of cancer cells and were considered to receive their nutrition by means of direct diffusion from the surrounding hepatic sinusoids.

The second pattern of hepatic metastases (mean diameter, approximately 290 µm) showed ICAM-1–positive intratumoral residual hepatic sinusoids that connected with the surrounding hepatic sinusoids. At this stage, with growth of the metastases, the neighboring hepatic sinusoids were compressed, were infiltrated, and were encircled, so the portal blood flow directly perfused them through the surrounding hepatic sinusoids. However, a few CD34-positive intratumoral microvessels also appeared at this stage, and this appearance indicated the beginning of tumor angiogenesis. ICAM-1–positive intratumoral microvessels also have been found in a clinical study about hepatic metastases, although the number, distribution, and origin of these microvessels were not explored in detail (22).

The third pattern showed hepatic metastases (mean diameter, approximately 520 µm), with convoluted intratumoral microvessels having decreased afferent microvessels and a blood supply from both the portal vein and the hepatic artery, as revealed with the study we performed with injection of fluorescent microspheres. Intratumoral CD34-positive microvessels and {alpha}-SMA–positive arterioles increased in this stage, and this finding indicated predominant tumor angiogenesis. The generation of new blood vessels, which are CD34 positive, usually involves sprouting angiogenesis from an existing vascular bed during various physiologic or pathologic conditions, which include malignancy (23). As for the angiogenesis of hepatic metastases, it is considered to be derived from the neighboring or entrapped hepatic sinusoids, arterioles in portal tracts, and isolated arterioles (2,3,9,24,25), all of which widely exist in mammalian livers (26,27). Results of our serial observation and findings in other research studies have suggested that some residual hepatic sinusoids in the tumor take part in the tumor-sprouting angiogenesis process (28), and the phenotype of the endothelium that lines the residual sinusoids may change during the tumor angiogenesis process (29,30). The fact that the CD34-positive microvessels, which were immunohistochemically negative on the normal hepatic sinusoidal endothelium, appeared and gradually became predominant in the tumor in accordance with the disappearance of intratumoral residual hepatic sinusoids strongly indicated this process.

In addition, the capillary network formation from sprouting angiogenesis is considered to be made by anastomoses between branches from arterioles and venules by a complicated guiding system (31). As a result, many irregularly dilated CD34 blood spaces appear in the tumor because of the immature wall of newly formed vessels (32) and the high pressure of hepatic artery perfusion. In contrast, the ICAM-1–positive residual sinusoids are distributed at the edge of hepatic metastases, communicating intratumoral microvessels, and neighboring sinusoids. Because of these communications, after high-pressure arterial blood inflow into the tumor, the direction of the blood flow in some intratumoral microvessels reverses to outflow to the surrounding hepatic sinusoids. The results obtained by using MVD and afferent MVD analyses also support this thinking. Blood perfusion derived from neighboring sinusoids also was observed at this stage, and the distribution of the intratumoral microvessels and blood flow in them was heterogeneous.

The fourth pattern was hepatic metastases (mean diameter, >2000 µm) with convoluted intratumoral microvessels that showed an exclusively arterial blood supply to them. Intratumoral microvessels were more regular and spatially homogeneous, probably caused by remodeling of the neovasculature and stabilization of the newly formed blood vessels mediated by recruitment of periendothelial support cells (32). According to the results of the study with injection of microspheres, the blood supply into the tumor was considered to be exclusively from the artery, and the blood in the tumor drained into the surrounding hepatic sinusoids. These findings are similar to the clinical observation reported by Terayama et al (20). In that study, findings at single-level dynamic CT during hepatic arteriography indicated that almost all of the blood flow in the metastases drained into the surrounding hepatic sinusoids through the communications with the intratumoral blood sinusoids.

To evaluate the arterial blood supply to hepatic lesions, {alpha}-SMA has been used as a special marker for intratumoral arterioles in many studies (17,18), although it was also immunohistochemically positive for many perisinusoidal stromal cells. However, these cells could be identified because they do not appear as a thick wall around a vessel lumen (33). In our study, the {alpha}-SMA–positive arterioles appeared in larger tumors that were larger than 500 µm in diameter, and all of these metastases showed pattern 3 at the immunohistochemical study and pattern III or IV at in vivo microscopy. These findings also strongly indicate that there was increased arterial blood supply in these large tumors.

Findings in previous studies about the perfusion of hepatic metastases in the early stage have caused much controversy in regard to the origin of the blood supply, namely, whether the blood supply is derived from the portal vein or the hepatic artery (2,3). The contention that both vessels can supply the hepatic metastases has been widely accepted (4,7,24); the hepatic artery plays a predominant role in supplying large metastatic tumors, and portal blood flow is important in small metastases (34,35). To explain this phenomenon, researchers have proposed various models (7,25,36). However, according to our observation, the blood perfusion of hepatic metastases and the development of intratumoral microvessels were continually changing phenomena in accordance with the growth of the metastases. To our knowledge, our study is the first to describe the transition from portal blood supply to arterial blood supply during development of hepatic metastases in vivo by taking into account the in vivo natural microcirculation, tumor angiogenesis, and interaction with the surrounding hepatic sinusoids. Our results suggest that the intratumoral microvessels might communicate with both the arterioles and hepatic sinusoids directly, with the blood flow direction in them being determined by the pressure changes occurring in these vessels. At the beginning of angiogenesis, arterioles were rare in the metastases, and there might be fewer communications between the intratumoral microvessels and these arterioles. On the other hand, residual hepatic sinusoids were common in the metastases at this stage. Therefore, the metastases might derive considerable blood perfusion from the portal vein through the hepatic sinusoids. With the growth of metastases, the number of intratumoral arterioles increased, and the communications between the intratumoral microvessels and these arterioles might also have increased. As a result, the arterial blood perfusion with high pressure became dominant, and it drained into the surrounding hepatic sinusoids through the communications with the intratumoral microvessels.

The high-pressure tumor blood flow entering the surrounding hepatic sinusoids could raise the resistance of the intrahepatic vascular bed and induce portal hypertension in tumor-bearing liver. This may account for a number of previously described observations: The dense vascular network surrounds the hepatic metastases that are being formed by dilated portal venules and sinusoids but not by hepatic arterioles (37); portal blood flow decreases and mean transit time increases in occult hepatic metastases detected with quantitative CT (5); and well-known peritumoral rim enhancement of overt hepatic metastases is present on CT and magnetic resonance images (38,39).

Our study had limitations. First, blood flow in arterioles could not be observed directly with this method because of the depth and small diameter of the arterioles. Second, the purpose of our study was not to compare the findings of in vivo microscopy with those of the immunohistochemical study for the same tumor because it was technically difficult to identify metastases one by one with in vivo microscopy. Third, the stepwise morphologic and hemodynamic changes of intratumoral microvessels may be specific for this cancer cell type; thus, different cancer cell lines that possess different metastastic properties (8) should be evaluated. Fourth, for the in vivo study, the mice were too small to perfectly simulate the clinical hepatic hemodynamic changes that occur in various physiologic or pathologic conditions because of the short circulation time of the portal system. Therefore, other larger animals should be investigated as well. Fifth, we did not evaluate data dependency caused by multiple tumors, and this factor could potentially affect the statistical results.

Practical application: The combined findings of in vivo microscopy and immunohistochemical analysis in our study strongly suggested that stepwise microcirculatory and morphologic changes of intratumoral microvessels occur during the establishment of hepatic colorectal cancer metastases from a minute focus of cancer cells to a macroscopic focus. Recognition of this phenomenon may be useful for improving image diagnosis, transcatheter therapy, and antiangiogenic therapy for hepatic metastases, and further knowledge of tumor vascularization would help in the evaluation of therapy with antiangiogenic drugs and the antivascular response of these drugs.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   FOOTNOTES
 

Abbreviations: {alpha}-SMA = alpha smooth muscle actin • ICAM-1 = intracellular adhesion molecule • MVD = microvessel density

2 Current address: Department of Radiology, First Affiliated Hospital of China Medical University, Shenyang, China. Back

Authors stated no financial relationship to disclose.

See also Science to Practice in this issue.

Author contributions: Guarantors of integrity of entire study, Y.L., O.M.; study concepts/study design or data acquisition or data analysis/interpretation, Y.L., O.M.; manuscript drafting or manuscript revision for important intellectual content, Y.L., O.M.; manuscript final version approval, Y.L., O.M.; literature research, Y.L., O.M.; experimental studies, Y.L.; statistical analysis, Y.L.; and manuscript editing, Y.L., O.M.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Baker ME, Pelley R. Hepatic metastases: basic principles and implications for radiologists. Radiology 1995;197:329–337.[Free Full Text]
  2. Gonda T, Ishida H, Yoshinaga K, Sugihara K. Microvasculature of small liver metastases in rats. J Surg Res 2000;94:43–48.[CrossRef][Medline]
  3. Kuruppu D, Christophi C, O'Brien PE. Microvascular architecture of hepatic metastases in a mouse model. HPB Surg 1997;10:149–158.[Medline]
  4. Casillas S, Dietz DW, Brand MI, Jones SC, Vladisavljevic A, Milsom JW. Perfusion to colorectal cancer liver metastases is not uniform and depends on tumor location and feeding vessel. J Surg Res 1997;67:179–185.[CrossRef][Medline]
  5. Cuenod C, Leconte I, Siauve N, et al. Early changes in liver perfusion caused by occult metastases in rats: detection with quantitative CT. Radiology 2001;218:556–561.[Abstract/Free Full Text]
  6. Yarmenitis SD, Kalogeropoulou CP, Hatjikondi O, et al. An experimental approach of the Doppler perfusion index of the liver in detecting occult hepatic metastases: histological findings related to the hemodynamic measurements in Wistar rats. Eur Radiol 2000;10:417–424.[CrossRef][Medline]
  7. Kan Z, Ivancev K, Lunderquist A, et al. In vivo microscopy of hepatic tumors in animal models: a dynamic investigation of blood supply to hepatic metastases. Radiology 1993;187:621–626.[Abstract/Free Full Text]
  8. Kruskal JB, Thomas P, Kane RA, Goldberg SN. Hepatic perfusion changes in mice livers with developing colorectal cancer metastases. Radiology 2004;231:482–490.[Abstract/Free Full Text]
  9. Fukumura D, Yuan F, Monsky WL, Chen Y, Jain RK. Effect of host microenvironment on the microcirculation of human colon adenocarcinoma. Am J Pathol 1997;151:679–688.[Abstract]
  10. Kruskal JB, Thomas P, Nasser I, Cay O, Kane RA. Hepatic colon cancer metastases in mice: dynamic in vivo correlation with hypoechoic rims visible at US. Radiology 2000;215:852–857.[Abstract/Free Full Text]
  11. Kitakata H, Nemoto-Sasaki Y, Takahashi Y, Kondo T, Mai M, Mukaida N. Essential roles of tumor necrosis factor receptor p55 in liver metastasis of intrasplenic administration of colon 26 cells. Cancer Res 2002;62:6682–6687.[Abstract/Free Full Text]
  12. Liu Y, Matsui O. Collaterals through hepatic sinusoids after embolization of terminal portal venules: an in vivo study on mice. Hepatol Res 2005;31:36–42.[CrossRef][Medline]
  13. Cui S, Hano H, Sakata A, et al. Enhanced CD34 expression of sinusoid-like vascular endothelial cells in hepatocellular carcinoma. Pathol Int 1996;46:751–756.[Medline]
  14. Kimura H, Nakajima T, Kagawa K, et al. Angiogenesis in hepatocellular carcinoma as evaluated by CD34 immunohistochemistry. Liver 1998;18:14–19.[Medline]
  15. Couvelard A, Scoazec JY, Dauge MC, Bringuier AF, Potet F, Feldmann G. Structural and functional differentiation of sinusoidal endothelial cells during liver organogenesis in humans. Blood 1996;87:4568–4580.[Abstract/Free Full Text]
  16. Medina J, Sanz-Cameno P, Garcia-Buey L, Martin-Vilchez S, Lopez-Cabrera M, Moreno-Otero R. Evidence of angiogenesis in primary biliary cirrhosis: an immunohistochemical descriptive study. J Hepatol 2005;42:124–131.[Medline]
  17. Himeno H, Enzan H, Saibara T, Onishi S, Yamamoto Y. Hitherto unrecognized arterioles within hepatocellular carcinoma. J Pathol 1994;174:217–222.[CrossRef][Medline]
  18. Morinaga S, Imada T, Shimizu A, et al. Angiogenesis in hepatocellular carcinoma as evaluated by alpha smooth muscle actin immunohistochemistry. Hepatogastroenterology 2001;48:224–228.[Medline]
  19. Matsui O, Takashima T, Kadoya M, et al. Liver metastases from colorectal cancers: detection with CT during arterial portography. Radiology 1987;165:65–69.[Abstract/Free Full Text]
  20. Terayama N, Matsui O, Ueda K, et al. Peritumoral rim enhancement of liver metastasis: hemodynamics observed on single-level dynamic CT during hepatic arteriography and histopathologic correlation. J Comput Assist Tomogr 2002;26:975–980.[CrossRef][Medline]
  21. Lin G, Lunderquist A, Hagerstrand I, Boijsen E. Postmortem examination of the blood supply and vascular pattern of small liver metastases in man. Surgery 1984;96:517–526.[Medline]
  22. Gulubova MV. Expression of cell adhesion molecules, their ligands and tumour necrosis factor alpha in the liver of patients with metastatic gastrointestinal carcinomas. Histochem J 2002;34:67–77.[CrossRef][Medline]
  23. Risau W. Mechanisms of angiogenesis. Nature 1997;386:671–674.[CrossRef][Medline]
  24. Terayama N, Terada T, Nakanuma Y. A morphometric and immunohistochemical study on angiogenesis of human metastatic carcinomas of the liver. Hepatology 1996;24:816–819.[CrossRef][Medline]
  25. Ackerman NB. The blood supply of experimental liver metastases. IV. Changes in vascularity with increasing tumor growth. Surgery 1974;75:589–596.
  26. Ekataksin W. The isolated artery: an intrahepatic arterial pathway that can bypass the lobular parenchyma in mammalian livers. Hepatology 2000;31:269–279.[CrossRef][Medline]
  27. Yamamoto K, Sherman I, Phillips MJ, Fisher MM. Three-dimensional observations of the hepatic arterial terminations in rat, hamster and human liver by scanning electron microscopy of microvascular casts. Hepatology 1985;5:452–456.[CrossRef][Medline]
  28. Gervaz P, Scholl B, Mainguene C, Poitry S, Gillet M, Wexner S. Angiogenesis of liver metastases: role of sinusoidal endothelial cells. Dis Colon Rectum 2000;43:980–986.[CrossRef][Medline]
  29. Sakamoto M, Ino Y, Fujii T, Hirohashi S. Phenotype changes in tumor vessels associated with the progression of hepatocellular carcinoma. Jpn J Clin Oncol 1993;23:98–104.[Abstract/Free Full Text]
  30. Hattori M, Fukuda Y, Imoto M, Koyama Y, Nakano I, Urano F. Histochemical properties of vascular and sinusoidal endothelial cells in liver diseases. Gastroenterol Jpn 1991;26:336–343.[Medline]
  31. Fuller T, Korff T, Kilian A, Dandekar G, Augustin HG. Forward EphB4 signaling in endothelial cells controls cellular repulsion and segregation from ephrinB2 positive cells. J Cell Sci 2003;116:2461–2470.[Abstract/Free Full Text]
  32. Asahara T, Chen D, Takahashi T, et al. Tie2 receptor ligands, angiopoietin-1 and angiopoietin-2, modulate VEGF-induced postnatal neovascularization. Circ Res 1998;83:233–240.[Abstract/Free Full Text]
  33. Enzan H, Himeno H, Iwamura S, et al. Alpha-smooth muscle actin-positive perisinusoidal stromal cells in human hepatocellular carcinoma. Hepatology 1994;19:895–903.[CrossRef][Medline]
  34. Ackerman NB, Lien WM, Kondi ES, Silverman NA. The blood supply of experimental liver metastases. I. The distribution of hepatic artery and portal vein blood to "small" and "large" tumors. Surgery 1969;66:1067–1072.
  35. Conway JG, Popp JA, Ji S, Thurman RG. Effect of size on portal circulation of hepatic nodules from carcinogen-treated rats. Cancer Res 1983;43:3374–3378.[Abstract/Free Full Text]
  36. Ackerman NB. Experimental studies on the role of the portal circulation in hepatic tumor vascularity. Cancer 1986;58:1653–1657.[CrossRef][Medline]
  37. Dingemans KP. B16 metastases in mouse liver and lung. II. Morphology. Invasion Metastasis 1988;8:87–102.
  38. Irie T, Tsushima Y, Terahata S, Hatsuse K, Kusano S. Rim enhancement in colorectal metastases at CT during infusion hepatic arteriography: does it represent liver parenchyma or live tumor cell zone? Acta Radiol 1997;38:416–421.[Medline]
  39. Semelka RC, Hussain SM, Marcos HB, Woosley JT. Perilesional enhancement of hepatic metastases: correlation between MR imaging and histopathologic findings—initial observations. Radiology 2000;215:89–94.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 RadiologyHome page
T. S. Koh, C. H. Thng, P. S. Lee, S. Hartono, H. Rumpel, B. C. Goh, and S. Bisdas
Hepatic Metastases: In Vivo Assessment of Perfusion Parameters at Dynamic Contrast-enhanced MR Imaging with Dual-Input Two-Compartment Tracer Kinetics Model
Radiology, August 11, 2008; (2008) 2483071958.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2432060341v1
243/2/386    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Liu, Y.
Right arrow Articles by Matsui, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Liu, Y.
Right arrow Articles by Matsui, O.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE