- Email: [email protected]
Anti-angiogenic Therapy on Hepatocellular Carcinoma Development and Progression
Journal of Surgical Research 158, 69–76 (2010) doi:10.1016/j.jss.2008.09.015
Anti-angiogenic Therapy on Hepatocellular Carcinoma Development and Progression Yuji Ishii, M.D., Ph.D.,1 Taro Sakamoto, M.D., Rysuke Ito, M.D., and Katsuhiko Yanaga, M.D., Ph.D. Department of Surgery, Jikei University School of Medicine, Tokyo, Japan Submitted for publication June 3, 2008
Key Words: hepatocellular carcinoma; angiogenesis; anti-angiogenic therapy; tumor dormancy; chemoprevention; IL-8.
Background. The hypervascular nature of hepatocellular carcinoma (HCC) has led a number of investigators to explore the potential of anti-angiogenic therapy for this cancer. We aimed to investigate the efficacy of anti-angiogenetic therapy for the prevention and treatment of HCC. Methods. An experimental model of HCC induced by diethylnitrosamine (DEN) was used in this study as a model resembling human HCC. Because endothelial cells play a central role in angiogenesis, FR-118487 (FR) with a direct effect on endothelial cells was selected for use in this study. FR-treated rats were divided into three groups of five animals each according to the timing of FR treatment, i.e., at the initiation of exposure to DEN (group I), after 6 wk of DEN exposure (group II), and after 12 wk of DEN exposure (group III). Results. No macroscopic nodules were observed in group I. Moreover, no significant difference in the mean maximum diameter and number of HCC nodules were found between the control group at 12 wk and group III at 18 wk. Thus, FR was shown to have a chemopreventive effect in addition to causing dormancy of HCC. The results obtained, including IL-8 levels, reflected the effect on HCC of anti-angiogenic therapy with FR. Conclusion. Anti-angiogenic therapy may be effective for both chemoprevention and tumor dormancy in HCC associated with chronic liver disease. IL-8 may be useful as a marker of the efficacy of anti-angiogenic therapy for HCC. Ó 2010 Elsevier Inc. All rights reserved.
INTRODUCTION
Hepatocellular carcinoma (HCC) develops as a consequence of underlying liver disease, most commonly viral hepatitis. The therapeutic options for patients with HCC are still limited. Only HCC patients with well-preserved liver function or no macroscopic vascular invasion are considered for curative treatments, such as resection, liver transplantation, or local ablation. Systemic treatment of advanced HCC is not very effective due to the marked resistance of this tumor to chemotherapy and radiation. In addition, most patients with HCC are highly susceptible to toxic side effects of systemic therapy because of impaired liver function. HCC is a hypervascular tumor, and neovascularization plays an important role in its growth and progression. Because angiogenesis increases during the early phase of the development of HCC [1], it provides a target for novel therapeutic approaches. Most clinical studies of anti-angiogenic agents in patients with HCC have focused on the response of advanced or end-stage disease. In the present study, FR-118487 (FR), an inhibitor of endothelial cells (ECs) proliferation, was used for stepwise treatment from the stage of hepatocarcinogenesis to tumor growth in an animal model of HCC induced by diethylnitrosamine (DEN). The reason for using the DEN model in this experiment is that it resembles human hepatocarcinogenesis [2]. FR is a synthetic analog of fumagillin that was isolated from Scolecobasidum arenarism, and has
1 To whom correspondence and reprint requests should be addressed at Department of Surgery, Jikei University School of Medicine, 3-25-8 Nishi-Shinbashi, Minato-ku, Tokyo, 105-8461, Japan. E-mail: [email protected].
69
0022-4804/10 $36.00 Ó 2010 Elsevier Inc. All rights reserved.
70
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
been confirmed to suppress neovascularization through its inhibitory effect on EC proliferation both in vitro and in vivo [3, 4]. We aimed to investigate the efficacy of anti-angiogenic therapy for HCC and for preventing hepatocarcinogenesis. MATERIALS AND METHODS Experimental HCC model Five-week-old male Fisher rats (Japan Charles River, Co., Ltd., Yokohama, Japan) were given diethylnitrosamine (DEN; Tokyo Chemical Industry, Co., Ltd., Tokyo, Japan) dissolved in distilled water at 100 ppm ad libitum for 12 wk and then received tap water for the next 6 wk. Two or three animals per cage were housed in a room with the temperature controlled at 23 6 2 C. Solid rat chow (CE-2; Clea Japan, Inc., Tokyo, Japan) was provided ad libitum. The rats were handled in accordance with the National Institutes of Health Guidelines for the Care of laboratory Animals and these experiments were approved by the Animal Research Committee of our university.
FR-118487 Treatment Protocol FR was dissolved in propylene glycol before use. FR was administered intraperitoneally at a dose of 1.5 mg/kg/d for 6 wk by continuous infusion via an Alzet osmotic pump (model 2002; Alza Corporation, Cupertino, CA). FR-treated rats were divided into three groups of five animals each according to the timing of treatment initiation, i.e., at the initiation of exposure to DEN (group I), after 6 wk of DEN exposure (group II), and after 12 wk of DEN exposure (group III). These three groups were set up to examine both the chemopreventive and therapeutic effects of FR for HCC. Figure 1 summarizes the experimental protocol and design.
Evaluation of Efficacy The FR-treated groups were compared with control animals, which underwent sham operation without FR treatment. Control group was administered with an equal volume of propylene glycol without FR and same surgical procedure. For convenience, the control group of group II was dealt with as a control group of group I. The control animals were sacrificed and liver specimens and blood samples (from the bifurcation of the abdominal aorta) were obtained at 6, 8, 10, 12, and 18 wk after starting our experiment (each group, n ¼ 5). The effects of FR were evaluated at 12 wk in groups I and II. On the other hand, the
effects of FR were evaluated at 18 weeks in group III, which was the end of the experimental protocol (Fig. 1). Evaluations included assessment of the macroscopic findings (the mean maximum size and number of tumors), creation of hepatic vascular casts by infusion of resin, measurement of microvessel density (MVD) by immunohistological analysis, and measurement of a-fetoprotein (AFP) and IL-8 levels in the serum.
Hepatic Vascular Casts Hepatic vascular casts were obtained by using a modification of the method of Skinner et al. [5]. Briefly, each animal was anesthetized with pentobarbital sodium (60 mg/kg, intraperitoneally). The thoracic aorta was exposed above the diaphragm to prevent injury to the liver during the surgical procedure. To infuse the resin, cannulas were inserted into the aorta and the portal vein, and secured with ties. Blood was then washed out from the vascular system by infusing filtered 0.9% NaCl with heparin at 10,000 units/L. An acrylic resin solution was prepared by mixing 20 g of red-colored Mercox CL-2R (Okenshoji Co., Ltd., Tokyo, Japan) and 0.5 g of Catalyst-MA (Okenshoji Co.) in a 20-mL syringe, and then was infused via the aortic cannula at 200 mmHg. During this procedure, 0.9% NaCl was infused via the portal vein continuously. Subsequently, blue-colored Mercox CL-2B (Okenshoji Co.) mixed with Catalyst-MA was infused via the portal cannula at 20 mmHg. After these resins were allowed to fully polymerize in situ at room temperature, each liver was carefully excised and cut into half; one-half was used for histopathological examination, and the other half was used for examination of the vascular cast. The half of the liver employed to obtain a vascular cast was rinsed with water before being placed in a 20% KOH solution at 37 C to digest the hepatic tissues. Dried casts were mounted on aluminium stubs, sputter-coated with gold (96%)-palladium (4%), and examined under a scanning electron microscope (SEM) [6].
Measurement of AFP and IL-8 Serum AFP was determined by a sandwich enzyme-linked immunosorbent assay (ELISA) (HOPE Laboratories Co., Ltd., Belmont, CA). Serum IL-8 was determined with a Rat GRO/CINC-1 Assay Kit (Immuno-Biological Laboratories Co., Ltd., Takasaki, Japan).
Immunohistochemical Analysis of the Microvessel Density Before performing procedure of hepatic vascular casts, the maximal nodules (HCC) on the liver surface in each group (n ¼ 5) were used for immunohistochemical analysis. The streptavidin-biotin (SAB) technique was used for immunohistochemical staining [7, 8]. Liver tissues were fixed with Methacarn solution (methanol:chloroform:glacial
FIG. 1. Experimental protocol and design: Rats were given DEN dissolved in distilled water at 100 ppm ad libitum for 12 wk and then received tap water for the next 6 wk. FR-118487 (FR) was administered intraperitoneally for 6 wk. FR-treated rats were divided into three groups (group I, group II, and group III). HCC: Hepatocellular carcinoma, DEN: Dietylnitrosamine, W: weeks.
ISHII ET AL.: ANTI-ANGIOGENIC THERAPY ON HCC acetic acid ¼ 6:3:1), embedded in paraffin, cut into 3-mm sections, and deparaffinized with a graded xylene series. Endogenous peroxidase was blocked by adding 0.3% H2O2 in methanol. Expression of CD31 (Novacastra Laboratories Ltd., Newcastle-upon-Tyne, UK) was identified by immunohistochemistry to evaluate tumor vessels [9]. Four fields in the cancerous region were observed. After these fields were identified at a magnification of 3200 using a Zeiss Axiovision 2.05 (Hallberg Moos, Germany), the results were analyzed with NIH Image software (NIH, Bethesda, MD) [10]. Briefly, we measured the percentage of 3.3’-diaminobenzidine (DAB)-stained microvessels in the unit frame and calculated the mean value as the MVD [11].
Statistical Analysis The mean and standard deviation were calculated for each of the parameters. To assess the statistical significance of intergroup differences for quantitative data, Bonferroni’s multiple comparison test was used after one-way ANOVA. All calculations were performed with Stat View-J statistical software (SAS Institute, Cary, NC) and P < 0.05 was considered significant.
RESULTS Hepatocarcinogenesis
The process of hepatocarcinogenesis in this DEN model was followed for up to 18 wk. The macroscopic changes included hyperplastic nodules at 6 wk and some whitish nodules at 8 wk, while nodular lesions suggestive of HCC usually appeared around 10 wk and increased up to 18 wk. Characteristic histological changes included portal–portal fibrous bridging at 6 wk, and basophilic hyperplastic nodules with cellular atypia as well as the development of well differentiated HCCs at 8 wk. The percentage of HCC nodules increased from 10 wk to 12 wk. Subsequently, the percentage of moderately to poorly differentiated HCCs gradually increased up to 18 wk (data not shown). Macroscopic Findings: Red Nodules
Red nodules detected after infusing resin were diagnosed as HCCs histopathologically. Therefore, the mean maximum diameter and number of these red nodules were compared between the control groups and the FR-treated groups. Macroscopically, red nodules were observed from 8 wk, and showed an increase in size and number at 12 wk. There was an increase of parenchymal occupation at 18 wk (Fig. 2A). On the other hand, no macroscopic nodular changes were observed in group I at 12 wk (Fig. 2B). Group II showed a significant reduction in the mean maximum diameter and number of red nodules from 9.2 6 1.30 mm and 9.6 6 1.14, respectively, in the control group at 12 weeks to 4.4 6 1.14 mm and 2.6 6 0.09, respectively (both P < 0.01). Likewise, group III showed a significant reduction in the mean maximum diameter and number of red nodules from 18.2 6 2.17 mm and 15.2 6 2.17, respectively, in the control group at 18 wk to 10.6 6
71
2.59 mm and 7.2 6 1.48, respectively (both P < 0.01). Moreover, no significant differences of the mean maximum diameter and number of red nodules were found between the control group at 12 wk and group III at 18 wk (Table 1). AFP and IL-8
Figure 3 shows serum AFP levels in the control and FR-treated groups. The serum AFP level of agematched normal animals without DEN or FR administration (measured simultaneously) was always less than 30 ng/mL. The serum AFP level of the control group began to increase from 6 wk and rose markedly by 18 wk with hepatocarcinogenesis and tumor growth. There was a wide difference between the serum AFP levels of normal animals and the control or FR-treated groups. The serum AFP level of group II was 9,748 6 2997 ng/mL at 12 wk, which was significantly lower than the value of 21,236 6 4024 ng/mL in the control group (P < 0.01). Furthermore, the AFP level of group I (1355 6 510 ng/mL) was significantly lower than that of either the control group or group II (both P < 0.01). At 18 wk, the serum AFP level was 89,660 6 24,578 ng/mL in group III, which was significantly lower than the value of 523,800 6 111,442 ng/mL in the control group (P < 0.01). Figure 4 shows serum IL-8 levels in the control and FR-treated groups. The serum IL-8 level of agematched normal animals without DEN or FR administration (measured simultaneously) was 84.88 6 18.57 pg/mL (n ¼ 5). In the control group, IL-8 was already increased at 6 wk and rose further by 18 wk. The serum IL-8 level of group II was 218.14 6 40.17 pg/mL at 12 wk, but was significantly lower than the value of 316.43 6 31.80 pg/mL in the control group (P < 0.01). The IL-8 level of group I (146.92 6 76.70 pg/mL) was significantly lower than that of either the control group or group II (both P < 0.01). Furthermore, there was no significant difference between group I and normal animals (P ¼ 0.078). On the other hand, there was a significant difference of the serum IL-8 level between group I and the control group 6 wk (P < 0.05). At 18 wk, the serum IL-8 level was 286.0 6 69.70 pg/mL in group III, which was significantly lower than the value of 358.57 6 40.18 pg/mL in the control group (P < 0.05). There was no significant difference of IL-8 between group III and the control group 12 weeks (P ¼ 0.252). Hepatic Vascular Casts
Scanning electron microscopy of vascular casts from normal rat livers revealed a typical sinusoidal network that converged on the central vein, and the sinusoids formed a continuous network with a regular diameter.
72
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
FIG. 2. (A) Macroscopic findings after infusion of resin: Multiple red nodules (diagnosed as hepatocellular carcinoma histopathologically) in the control group at 18 wk. (B) No distinct nodules are observed in group I (before infusing resin).
A large portal vein, a small hepatic artery, and sometimes a peribiliary plexus were observed in the portal tract (Fig. 5A, B). The hyperplastic nodules that appeared around 6 wk were filled with blue resin similar to the surrounding tissues. The sinusoids of these nodules showed irregular, narrowing, and were roundish (Fig. 5C). The red nodules were diagnosed as HCCs histopathologically and their sinusoids showed a meshlike pattern at 12 wk (Fig. 5D, E), with the irregularity increasing further at 18 wk (Fig. 5F). Tumor angiogenesis was conspicuous at the edges of the red nodules (Fig. 5G). Figure 4H shows the structure of a red nodule from group III. Vascular casts of red nodules showed a decrease of irregularity at the sinusoidal level and a sparse structure in groups II and III. Tumor angiogenesis at the edges of the nodules was also inhibited. Furthermore, the normal sinusoidal structure was maintained in group I (Fig. 5I). Microvessel Density
MVD in tumor tissues was stained by CD31 immunohistochemistry using DAB (Fig. 6). Expression of CD31 in the sinusoids of experimental HCCs induced by DEN increased gradually from 6 wk to 18 wk during the process of carcinogenesis and tumor growth (Fig. 7). The MVD of normal liver tissue was less than 1.2%. FR treatment significantly reduced the MVD, as demonstrated by computer-assisted image analysis of sections immunostained for CD31. Figure 7 shows a comparison between the respective control groups and the FRtreated groups. Compared with the control group at 12 wk, the MVD was reduced by 35% in group II and about 49% in group I (both P < 0.01). Furthermore, there was significant difference of MVD between group I and group II (P < 0.05). Likewise, compared with the control group at 18 wk, MVD was reduced by 40% in
group III (P < 0.01). There was no significant difference of MVD between group III and the control group 12 wk (P ¼ 0.0143). Adverse Effects: Body Weight
At 12 wk, the mean body weight was 145.20 6 22.67 g in group II, which was significantly lower than that in the control group (226.40 6 15.65 g, P < 0.01). There was no significant difference of mean body weight between group I and group II. At 18 wk, the mean body weight was 183.40 6 15.39 g in group III, which was lower than that in the control group (234.80 6 14.13 g, P < 0.01). There were no significant differences in the consumption of DEN between the control groups and the FR-treated groups. DISCUSSION
Curative treatments currently available for HCC include surgical resection, liver transplantation, and percutaneous ablation. However, many patients have TABLE 1 Comparison of the Mean Maximum Diameter and the Mean Number of Red Nodules Red nodules Maximal diameter (mm) Control group (12 wk) Group I Group II Control group (18 wk) Group III a
P < 0.01. P < 0.01.
b
9.2 6 1.30a – 4.4 6 1.14a 18.2 6 2.17b 10.6 6 2.59b
Number 9.6 6 1.14a – 2.6 6 0.09a 15.2 6 2.17b 7.2 6 1.48b
ISHII ET AL.: ANTI-ANGIOGENIC THERAPY ON HCC
73
FIG. 3. Serum AFP levels in the control and FR-treated groups: (A) The serum AFP level of the control group rose at 6 wk and increased further up to 18 wk. The serum AFP levels of group I or II were significantly lower at 12 wk than that of the control group. (B) The serum AFP level of group III was significantly lower than that of the control group. a, b, cP <0.01, dP <0.01. Cont.: Control, W: weeks.
advanced HCC for which these treatments are not indicated. These patients are candidates for palliative treatment, which includes transarterial embolization, hormone therapy, interferon therapy, and chemotherapy. The response rate achieved with systemic chemotherapy is generally less than 30% [12], and patients often tolerate such a treatment poorly. Accordingly, we need new treatment strategies for HCC. In 1971, Folkman suggested the use of anti-angiogenic agents to inhibit tumors, a proposal which led to a new modality of anticancer treatment [13]. HCC is a typical hypervascular tumor; many angiogenic factors have been studied in this cancer, and several anti-angiogenic agents have been tested in animals and patients [14–16]. Angiogenesis is regulated by a balance between stimulators and inhibitors under normal conditions [17]. Some regulators of angiogenesis specifically target ECs, such as vascular endothelial cell growth factor (VEGF), angiopoietins, and endostatin, while others have multiple biological functions in addition to their role in angiogenesis, such as various chemokines that are also involved in the inflammatory response [14]. The relation between inflammation and cancer has attracted increasing attention because of the observation that chronic inflammation is usually associated with cancer [18]. A model of HCC induced by DEN was used in this study because it resembles human HCC [2]. A review of anti-angiogenic therapy for HCC showed that examination of anti-angiogenic treatment during hepatocarcinogenesis has not been performed before [14]. Therefore, this was taken into consideration in our study. Three-dimensional changes of the vasculature during hepatocarcinogenesis induced by DEN were studied
by SEM observation of vascular casts created by infusing red resin via the hepatic artery and blue resin via the portal vein. Nodules that appeared at 6 wk were filled with blue resin similar to the surrounding sinusoids, and the vascular casts of these nodules were roundish and irregular. Thus, changes of the vascular
FIG. 4. Serum IL-8 levels of the control and FR-treated groups: IL-8 was already increased at 6 wk, and rose further up to 18 wk. The serum IL-8 levels of groups I or II were significantly lower at 12 weeks than that of the control group. The serum IL-8 level of the group III was significantly lower than that of the control group. On the other hand, there was no significant difference between group III and the control 12 wk. There was a significant difference of serum IL-8 between group I and the control group 6 weeks. On the other hand, there was no significant difference between group I and normal animals. a, b, cP < 0.01, d, fP < 0.05, eP ¼ 0.252, gP ¼ 0.078. Cont.: Control, W: weeks.
74
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
FIG. 5. Scanning electron microscopy of vascular casts: (A) Normal liver. HA: hepatic arteriole, PV: portal venule. (B) High power view of a normal sinusoid. CHV: central hepatic vein. (C) Hyperplastic nodules at 6 wk. (D) Blood sinusoids in the control group at 12 wk. (E) High power view of blood sinusoids in the control group at 12 wk. (F) Blood sinusoids in the control group at 18 wk. (G) Angiogenesis at the border of a red nodule. (H) Sinusoidal architecture in group III. (I) Sinusoidal architecture in group I.
architecture were already observed at this time. The red nodules seen at 12 wk had incompletely anastomosing sinusoids with a coral-like appearance. These findings are typical of so-called ‘‘blood sinuses’’ and the red nodules were shown to be HCCs histopathologically. Furthermore, angiogenesis was prominent at the edges of these nodules. Red nodules (HCCs) at 18 wk exclusively contained arterial red resin that outlined irregular, mesh-like, anastomosing sinusoids. The importance of angiogenesis in the process of hepatocarcinogenesis was clarified by this morphological study. Endothelial lining cells usually express CD31 and CD34, but these markers are rarely associated with normal sinusoidal ECs [16]. The so-called capillarization of tumor-associated vessels may either result from the development of new vessels through the process of tumor angiogenesis or from the alteration of pre-existing sinusoids in response to tumor-induced
changes of the microenvironment [19, 20]. We immunohistochemically examined the expression of CD31, which is an endothelial-specific marker. CD31 binds with PECAM-1, a cell–-cell adhesion molecule of the immunoglobulin superfamily expressed by most endothelial cells [21]. In the control group, MVD was increased from 6 wk, and increased further by 18 wk. Furthermore, IL-8 (an angiogenic factor) was elevated at 6 wk and increased further after that with the development of the HCCs. Our findings strongly suggest that angiogenesis is closely related to hepatocarcinogenesis and tumor growth. Because endothelial cells play a central role in angiogenesis, FR (which directly acts on endothelial cells) was selected for this study. FR is an angiogenesis inhibitor that has been confirmed to suppress neovascularization through its inhibitory effect on EC proliferation both in vitro and in vivo [3, 4]. FR was synthesized by
ISHII ET AL.: ANTI-ANGIOGENIC THERAPY ON HCC
75
FIG. 6. The assessment of microvessel density (MVD) in tumor tissues. MVD was stained by CD31 immunohistochemistry using diaminobenzidine (DAB) and the image was displayed on the monitor with a Zeiss Axiovision 2.05. (A) Control 12 wk (3200), (B) group II (3200).
modification of the fermentation products of the fungus Scolecobasidum arenarium at Fujisawa Pharmaceutical Co., Ltd. (Tsukuba, Japan) [3]. In addition to its effect on HCC, the optimum timing of FR administration was examined during the process of hepatocarcinogenesis and tumor growth. No HCCs developed in group I, suggesting the usefulness of FR as a chemopreventive agent. As data from group II showed, even if anti-angiogenic treatment is started before the development of HCC, it is impossible to control the tumor subsequently. The point of debate about chemoprevention with anti-angiogenic therapy
FIG. 7. Immunohistochemical analysis of microvessel density (MVD). FR-118487 significantly reduced the MVD. Compared with the control group at 12 wk, the MVD was reduced by 35% in group II and by 49% in group I. There was also a significant difference of MVD between group I and group II. Compared with the control group, MVD in group III was reduced by 40% at 18 wk. There was no significant difference of MVD between group II and the control group at 18 weeks. a, b, dP < 0.01, cP < 0.05, eP ¼ 0.0143.
is when to start such treatment for patients with chronic liver disease before the development of HCC. However, the results for groups I and II indicate that starting anti-angiogenic treatment as soon as possible is better for chemoprevention. The mean maximum size and number of red nodules in groups II and III were reduced compared with the control group. On the other hand, there were no significant differences in the maximum size and number of red nodules between the control group at 12 wk and group III at 18 wk. This suggests that FR was able to induce a dormant state of HCC. Serum AFP levels also reflected these effects of FR. Examination of hepatic vascular casts and comparison with the control group showed that there was less irregularity blood sinuses in groups II and III and avascular areas were conspicuous. Angiogenesis to create feeding vessels for the red nodules was also inhibited. In small hyperplastic nodules from group I, the sinusoidal architecture was closer to that of normal liver than in the other groups. Immunohistochemistry for CD31 showed that MVD was significantly lower in groups II and III than in the respective control groups. Measurement of IL-8 also supported the results of immunohistochemical examination. It has been reported that IL-8 induces chemotactic activity on ECs and active angiogenesis in vivo [22]. Potently angiogenic, IL-8 is constitutively produced by various tumor cell lines, including HCC lines [23]. Biological markers are needed to demonstrate the mechanisms of anti-angiogenic therapy and follow up its effects because conventional criteria for complete response (CR) or partial response (PR) fail to show the effects of such treatment. Such markers may include the serum concentrations of angiogenic factors. In this study, the serum level of IL-8 increased before the development of HCCs and rose further after the tumors appeared. In the FR-treated groups, IL-8 levels were lower compared with those in the control group. Our investigation of IL8 in HCC patients showed that it was elevated with
76
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
tumor development, and the elevation was considered to be important for hepatocarcinogenesis [24]. Accordingly, IL-8 may possibly serve as a marker for the effect of anti-angiogenic treatment on HCC (including chemoprevention). However, further studies are needed to establish anti-angiogenic therapy for chemoprevention using IL-8 as an index. The inhibition of weight gain is a major adverse effect of this therapy. The optimum dose, duration, and route of administration for clinical application still need to be determined. In regard to this, the continuous infusion of FR suppressed liner metastasis without any adverse effects [25]. FR-treated rats did not show side effects such as hair loss or infection that occur with conventional chemotherapy. In conclusion, FR markedly inhibited the development and progression of HCC through inhibition of angiogenesis in an animal model of HCC induced by DEN. Anti-angiogenic therapy may be effective for chemoprevention and tumor dormancy in patients who have HCC associated with chronic liver disease. Furthermore, IL-8 may be a useful marker for monitoring antiangiogenic treatment in HCC patients. ACKNOWLEDGMENTS The authors gratefully acknowledge Professor H. Hano, Department of Pathology, Jikei University School of Medicine, for excellent technical support.
REFERENCES 1. Li CY, Shan S, Huang Q, et al. Initial stages of tumor cellinduced angiogenesis: Evaluation via skin window chambers in rodent models. J Natl Cancer Inst 2000;92:143. 2. Masaka A. Sequential immunohistochemical study of early hepatocellular carcinoma in rat liver during carcinogenesis induced by diethylnitrosamine. St Mariana Med J (Japanese) 1992;20:472. 3. Otsuka T, Okawa T, Shibata T, et al. A new potent angiogenesis inhibitor, FR-118487. J Microbiol Biotechnol 1991;1:163. 4. Terano H, Shibata T, Otsuka T. Angiogenesis inhibitors of microbial origin. Drug Future 1993;18:239. 5. Skinner SA, Tutton PJM, O’Brien PE. Microvascular architecture of experimental colon tumors in the rat. Cancer Res 1990; 50:2411. 6. Tsuda H, Masui T, Inoue T. Three-dimensional observation of vasculature and blood supply during hepatocarcinogenesis in rats. Saishinigaku (Japanese) 1986;41:67. 7. Guesdon JL, Ternynck T, Avrames S. The use of avidin-biotin interaction in immunoenzymatic techniques. J Histochem Cytochem 1979;27:1131.
8. Wamke R, Levy R. Detection of T and B cell antigens with hybridoma monoclonal antibodies: A biotin-avidin-horseradish peroxidase method. J Histochem Cytochem 1980;28:771. 9. Yoshiji H, Kuriyama S, Ways DK, et al. Protein kinase C lies on the signaling pathway for vascular endothelial growth factormediated tumor development and angiogenesis. Cancer Res 1999;59:4413. 10. Kimura H, Nakajima T, Karawa K, et al. Angiogenesis in hepatocellular carcinoma as evaluated by CD34 immunohistochemistry. Liver 1998;18:14. 11. Moriguchi M, Nakajima T, Kimura H, et al. The copper chelator trientine has an anti-angiogenic effect against hepatocellular carcinoma, possibly through inhibition of interleukin-8 production. Int J Cancer 2002;102:445. 12. Thomas MB, Chadha R, Glover K, et al. Phase 2 study of erlotinib with unresectable hepatocellular carcinoma. Cancer 2007; 110:1059. 13. Folkman J. Tumor angiogenesis: Therapeutic implications. N Engl J Med 1971;285:1182. 14. Hui CS, Zaho YT. Angiogenesis in hepatocellular carcinoma: The retrospectives and perspectives. J cancer Res Clin Oncol 2004;130:307. 15. Zhong LZ, Zhi SL, Quan S. Effects of thalidomide on angiogenesis and tumor growth and metastasis of human hepatocellular carcinoma in nude mice. World J Gastroenterol 2005;11:216. 16. Schwartz M, Roayaie S, Knostadoulakis M. Strategies for the management of hepatocellular carcinoma. Nat Clin Pract Oncol 2007;4:424. 17. Glaser BM. Extracellular modulating factors and the control of intraocular neovascularization: An overview. Arch Opthamol 1988;106:603. 18. Lisa MC, Werb Z. Inflammation and cancer. Nature 2002; 420:860. 19. Nakamura S, Muro H, Suzuki S, et al. Immunohistochemical studies on endothelial cell phenotype in hepatocellular carcinoma. Hepatology 1997;26:407. 20. Frachon S, Gouysse G, Dumortier J, et al. Endothelial cell marker expression in dysplastic lesions of the liver: An immunohistochemical study. J Hepatol 2001;34:850. 21. Newman P, Berndt M, Gorski J, et al. PECAM-1 (CD31) cloning and relation to adhesion molecules of the immunoglobulin gene superfamily. Science 1990;247:1219. 22. Arenberg DA, Kunkel SL, Polverini PJ, et al. Inhibition of interleukin-8 reduces tumorigenesis of human non-small cell lung cancer in SCID mice. J Clin Invest 1996;97:2792. 23. Kubo F, Ueno S, Hiwatashi K, et al. Interleukin 8 in human hepatocellular carcinoma correlates with cancer cell invasion of vessels but not with tumor angiogenesis. Ann Surg Oncol 2005;12:800. 24. Ishii Y, Nakasato Y, Yanagisawa S, et al. Dose anti-angiogenic treatment make any contribution in cases of primary hepatocellular carcinoma? With special reference to chemoprevention, tumor dormancy and tumor shrinkage. Bioterapy (Japanese) 2002;16:396. 25. Ueoka M. Inhibitory effects of the angiogenesis inhibitor FR118487 on liver metastasis in a rabbit colon cancer model by continuous intraarterial infusion. Jpn J Gastroenterol Surg (Japanese) 1999;32:983.
Anti-angiogenic Therapy on Hepatocellular Carcinoma Development and Progression Yuji Ishii, M.D., Ph.D.,1 Taro Sakamoto, M.D., Rysuke Ito, M.D., and Katsuhiko Yanaga, M.D., Ph.D. Department of Surgery, Jikei University School of Medicine, Tokyo, Japan Submitted for publication June 3, 2008
Key Words: hepatocellular carcinoma; angiogenesis; anti-angiogenic therapy; tumor dormancy; chemoprevention; IL-8.
Background. The hypervascular nature of hepatocellular carcinoma (HCC) has led a number of investigators to explore the potential of anti-angiogenic therapy for this cancer. We aimed to investigate the efficacy of anti-angiogenetic therapy for the prevention and treatment of HCC. Methods. An experimental model of HCC induced by diethylnitrosamine (DEN) was used in this study as a model resembling human HCC. Because endothelial cells play a central role in angiogenesis, FR-118487 (FR) with a direct effect on endothelial cells was selected for use in this study. FR-treated rats were divided into three groups of five animals each according to the timing of FR treatment, i.e., at the initiation of exposure to DEN (group I), after 6 wk of DEN exposure (group II), and after 12 wk of DEN exposure (group III). Results. No macroscopic nodules were observed in group I. Moreover, no significant difference in the mean maximum diameter and number of HCC nodules were found between the control group at 12 wk and group III at 18 wk. Thus, FR was shown to have a chemopreventive effect in addition to causing dormancy of HCC. The results obtained, including IL-8 levels, reflected the effect on HCC of anti-angiogenic therapy with FR. Conclusion. Anti-angiogenic therapy may be effective for both chemoprevention and tumor dormancy in HCC associated with chronic liver disease. IL-8 may be useful as a marker of the efficacy of anti-angiogenic therapy for HCC. Ó 2010 Elsevier Inc. All rights reserved.
INTRODUCTION
Hepatocellular carcinoma (HCC) develops as a consequence of underlying liver disease, most commonly viral hepatitis. The therapeutic options for patients with HCC are still limited. Only HCC patients with well-preserved liver function or no macroscopic vascular invasion are considered for curative treatments, such as resection, liver transplantation, or local ablation. Systemic treatment of advanced HCC is not very effective due to the marked resistance of this tumor to chemotherapy and radiation. In addition, most patients with HCC are highly susceptible to toxic side effects of systemic therapy because of impaired liver function. HCC is a hypervascular tumor, and neovascularization plays an important role in its growth and progression. Because angiogenesis increases during the early phase of the development of HCC [1], it provides a target for novel therapeutic approaches. Most clinical studies of anti-angiogenic agents in patients with HCC have focused on the response of advanced or end-stage disease. In the present study, FR-118487 (FR), an inhibitor of endothelial cells (ECs) proliferation, was used for stepwise treatment from the stage of hepatocarcinogenesis to tumor growth in an animal model of HCC induced by diethylnitrosamine (DEN). The reason for using the DEN model in this experiment is that it resembles human hepatocarcinogenesis [2]. FR is a synthetic analog of fumagillin that was isolated from Scolecobasidum arenarism, and has
1 To whom correspondence and reprint requests should be addressed at Department of Surgery, Jikei University School of Medicine, 3-25-8 Nishi-Shinbashi, Minato-ku, Tokyo, 105-8461, Japan. E-mail: [email protected].
69
0022-4804/10 $36.00 Ó 2010 Elsevier Inc. All rights reserved.
70
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
been confirmed to suppress neovascularization through its inhibitory effect on EC proliferation both in vitro and in vivo [3, 4]. We aimed to investigate the efficacy of anti-angiogenic therapy for HCC and for preventing hepatocarcinogenesis. MATERIALS AND METHODS Experimental HCC model Five-week-old male Fisher rats (Japan Charles River, Co., Ltd., Yokohama, Japan) were given diethylnitrosamine (DEN; Tokyo Chemical Industry, Co., Ltd., Tokyo, Japan) dissolved in distilled water at 100 ppm ad libitum for 12 wk and then received tap water for the next 6 wk. Two or three animals per cage were housed in a room with the temperature controlled at 23 6 2 C. Solid rat chow (CE-2; Clea Japan, Inc., Tokyo, Japan) was provided ad libitum. The rats were handled in accordance with the National Institutes of Health Guidelines for the Care of laboratory Animals and these experiments were approved by the Animal Research Committee of our university.
FR-118487 Treatment Protocol FR was dissolved in propylene glycol before use. FR was administered intraperitoneally at a dose of 1.5 mg/kg/d for 6 wk by continuous infusion via an Alzet osmotic pump (model 2002; Alza Corporation, Cupertino, CA). FR-treated rats were divided into three groups of five animals each according to the timing of treatment initiation, i.e., at the initiation of exposure to DEN (group I), after 6 wk of DEN exposure (group II), and after 12 wk of DEN exposure (group III). These three groups were set up to examine both the chemopreventive and therapeutic effects of FR for HCC. Figure 1 summarizes the experimental protocol and design.
Evaluation of Efficacy The FR-treated groups were compared with control animals, which underwent sham operation without FR treatment. Control group was administered with an equal volume of propylene glycol without FR and same surgical procedure. For convenience, the control group of group II was dealt with as a control group of group I. The control animals were sacrificed and liver specimens and blood samples (from the bifurcation of the abdominal aorta) were obtained at 6, 8, 10, 12, and 18 wk after starting our experiment (each group, n ¼ 5). The effects of FR were evaluated at 12 wk in groups I and II. On the other hand, the
effects of FR were evaluated at 18 weeks in group III, which was the end of the experimental protocol (Fig. 1). Evaluations included assessment of the macroscopic findings (the mean maximum size and number of tumors), creation of hepatic vascular casts by infusion of resin, measurement of microvessel density (MVD) by immunohistological analysis, and measurement of a-fetoprotein (AFP) and IL-8 levels in the serum.
Hepatic Vascular Casts Hepatic vascular casts were obtained by using a modification of the method of Skinner et al. [5]. Briefly, each animal was anesthetized with pentobarbital sodium (60 mg/kg, intraperitoneally). The thoracic aorta was exposed above the diaphragm to prevent injury to the liver during the surgical procedure. To infuse the resin, cannulas were inserted into the aorta and the portal vein, and secured with ties. Blood was then washed out from the vascular system by infusing filtered 0.9% NaCl with heparin at 10,000 units/L. An acrylic resin solution was prepared by mixing 20 g of red-colored Mercox CL-2R (Okenshoji Co., Ltd., Tokyo, Japan) and 0.5 g of Catalyst-MA (Okenshoji Co.) in a 20-mL syringe, and then was infused via the aortic cannula at 200 mmHg. During this procedure, 0.9% NaCl was infused via the portal vein continuously. Subsequently, blue-colored Mercox CL-2B (Okenshoji Co.) mixed with Catalyst-MA was infused via the portal cannula at 20 mmHg. After these resins were allowed to fully polymerize in situ at room temperature, each liver was carefully excised and cut into half; one-half was used for histopathological examination, and the other half was used for examination of the vascular cast. The half of the liver employed to obtain a vascular cast was rinsed with water before being placed in a 20% KOH solution at 37 C to digest the hepatic tissues. Dried casts were mounted on aluminium stubs, sputter-coated with gold (96%)-palladium (4%), and examined under a scanning electron microscope (SEM) [6].
Measurement of AFP and IL-8 Serum AFP was determined by a sandwich enzyme-linked immunosorbent assay (ELISA) (HOPE Laboratories Co., Ltd., Belmont, CA). Serum IL-8 was determined with a Rat GRO/CINC-1 Assay Kit (Immuno-Biological Laboratories Co., Ltd., Takasaki, Japan).
Immunohistochemical Analysis of the Microvessel Density Before performing procedure of hepatic vascular casts, the maximal nodules (HCC) on the liver surface in each group (n ¼ 5) were used for immunohistochemical analysis. The streptavidin-biotin (SAB) technique was used for immunohistochemical staining [7, 8]. Liver tissues were fixed with Methacarn solution (methanol:chloroform:glacial
FIG. 1. Experimental protocol and design: Rats were given DEN dissolved in distilled water at 100 ppm ad libitum for 12 wk and then received tap water for the next 6 wk. FR-118487 (FR) was administered intraperitoneally for 6 wk. FR-treated rats were divided into three groups (group I, group II, and group III). HCC: Hepatocellular carcinoma, DEN: Dietylnitrosamine, W: weeks.
ISHII ET AL.: ANTI-ANGIOGENIC THERAPY ON HCC acetic acid ¼ 6:3:1), embedded in paraffin, cut into 3-mm sections, and deparaffinized with a graded xylene series. Endogenous peroxidase was blocked by adding 0.3% H2O2 in methanol. Expression of CD31 (Novacastra Laboratories Ltd., Newcastle-upon-Tyne, UK) was identified by immunohistochemistry to evaluate tumor vessels [9]. Four fields in the cancerous region were observed. After these fields were identified at a magnification of 3200 using a Zeiss Axiovision 2.05 (Hallberg Moos, Germany), the results were analyzed with NIH Image software (NIH, Bethesda, MD) [10]. Briefly, we measured the percentage of 3.3’-diaminobenzidine (DAB)-stained microvessels in the unit frame and calculated the mean value as the MVD [11].
Statistical Analysis The mean and standard deviation were calculated for each of the parameters. To assess the statistical significance of intergroup differences for quantitative data, Bonferroni’s multiple comparison test was used after one-way ANOVA. All calculations were performed with Stat View-J statistical software (SAS Institute, Cary, NC) and P < 0.05 was considered significant.
RESULTS Hepatocarcinogenesis
The process of hepatocarcinogenesis in this DEN model was followed for up to 18 wk. The macroscopic changes included hyperplastic nodules at 6 wk and some whitish nodules at 8 wk, while nodular lesions suggestive of HCC usually appeared around 10 wk and increased up to 18 wk. Characteristic histological changes included portal–portal fibrous bridging at 6 wk, and basophilic hyperplastic nodules with cellular atypia as well as the development of well differentiated HCCs at 8 wk. The percentage of HCC nodules increased from 10 wk to 12 wk. Subsequently, the percentage of moderately to poorly differentiated HCCs gradually increased up to 18 wk (data not shown). Macroscopic Findings: Red Nodules
Red nodules detected after infusing resin were diagnosed as HCCs histopathologically. Therefore, the mean maximum diameter and number of these red nodules were compared between the control groups and the FR-treated groups. Macroscopically, red nodules were observed from 8 wk, and showed an increase in size and number at 12 wk. There was an increase of parenchymal occupation at 18 wk (Fig. 2A). On the other hand, no macroscopic nodular changes were observed in group I at 12 wk (Fig. 2B). Group II showed a significant reduction in the mean maximum diameter and number of red nodules from 9.2 6 1.30 mm and 9.6 6 1.14, respectively, in the control group at 12 weeks to 4.4 6 1.14 mm and 2.6 6 0.09, respectively (both P < 0.01). Likewise, group III showed a significant reduction in the mean maximum diameter and number of red nodules from 18.2 6 2.17 mm and 15.2 6 2.17, respectively, in the control group at 18 wk to 10.6 6
71
2.59 mm and 7.2 6 1.48, respectively (both P < 0.01). Moreover, no significant differences of the mean maximum diameter and number of red nodules were found between the control group at 12 wk and group III at 18 wk (Table 1). AFP and IL-8
Figure 3 shows serum AFP levels in the control and FR-treated groups. The serum AFP level of agematched normal animals without DEN or FR administration (measured simultaneously) was always less than 30 ng/mL. The serum AFP level of the control group began to increase from 6 wk and rose markedly by 18 wk with hepatocarcinogenesis and tumor growth. There was a wide difference between the serum AFP levels of normal animals and the control or FR-treated groups. The serum AFP level of group II was 9,748 6 2997 ng/mL at 12 wk, which was significantly lower than the value of 21,236 6 4024 ng/mL in the control group (P < 0.01). Furthermore, the AFP level of group I (1355 6 510 ng/mL) was significantly lower than that of either the control group or group II (both P < 0.01). At 18 wk, the serum AFP level was 89,660 6 24,578 ng/mL in group III, which was significantly lower than the value of 523,800 6 111,442 ng/mL in the control group (P < 0.01). Figure 4 shows serum IL-8 levels in the control and FR-treated groups. The serum IL-8 level of agematched normal animals without DEN or FR administration (measured simultaneously) was 84.88 6 18.57 pg/mL (n ¼ 5). In the control group, IL-8 was already increased at 6 wk and rose further by 18 wk. The serum IL-8 level of group II was 218.14 6 40.17 pg/mL at 12 wk, but was significantly lower than the value of 316.43 6 31.80 pg/mL in the control group (P < 0.01). The IL-8 level of group I (146.92 6 76.70 pg/mL) was significantly lower than that of either the control group or group II (both P < 0.01). Furthermore, there was no significant difference between group I and normal animals (P ¼ 0.078). On the other hand, there was a significant difference of the serum IL-8 level between group I and the control group 6 wk (P < 0.05). At 18 wk, the serum IL-8 level was 286.0 6 69.70 pg/mL in group III, which was significantly lower than the value of 358.57 6 40.18 pg/mL in the control group (P < 0.05). There was no significant difference of IL-8 between group III and the control group 12 weeks (P ¼ 0.252). Hepatic Vascular Casts
Scanning electron microscopy of vascular casts from normal rat livers revealed a typical sinusoidal network that converged on the central vein, and the sinusoids formed a continuous network with a regular diameter.
72
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
FIG. 2. (A) Macroscopic findings after infusion of resin: Multiple red nodules (diagnosed as hepatocellular carcinoma histopathologically) in the control group at 18 wk. (B) No distinct nodules are observed in group I (before infusing resin).
A large portal vein, a small hepatic artery, and sometimes a peribiliary plexus were observed in the portal tract (Fig. 5A, B). The hyperplastic nodules that appeared around 6 wk were filled with blue resin similar to the surrounding tissues. The sinusoids of these nodules showed irregular, narrowing, and were roundish (Fig. 5C). The red nodules were diagnosed as HCCs histopathologically and their sinusoids showed a meshlike pattern at 12 wk (Fig. 5D, E), with the irregularity increasing further at 18 wk (Fig. 5F). Tumor angiogenesis was conspicuous at the edges of the red nodules (Fig. 5G). Figure 4H shows the structure of a red nodule from group III. Vascular casts of red nodules showed a decrease of irregularity at the sinusoidal level and a sparse structure in groups II and III. Tumor angiogenesis at the edges of the nodules was also inhibited. Furthermore, the normal sinusoidal structure was maintained in group I (Fig. 5I). Microvessel Density
MVD in tumor tissues was stained by CD31 immunohistochemistry using DAB (Fig. 6). Expression of CD31 in the sinusoids of experimental HCCs induced by DEN increased gradually from 6 wk to 18 wk during the process of carcinogenesis and tumor growth (Fig. 7). The MVD of normal liver tissue was less than 1.2%. FR treatment significantly reduced the MVD, as demonstrated by computer-assisted image analysis of sections immunostained for CD31. Figure 7 shows a comparison between the respective control groups and the FRtreated groups. Compared with the control group at 12 wk, the MVD was reduced by 35% in group II and about 49% in group I (both P < 0.01). Furthermore, there was significant difference of MVD between group I and group II (P < 0.05). Likewise, compared with the control group at 18 wk, MVD was reduced by 40% in
group III (P < 0.01). There was no significant difference of MVD between group III and the control group 12 wk (P ¼ 0.0143). Adverse Effects: Body Weight
At 12 wk, the mean body weight was 145.20 6 22.67 g in group II, which was significantly lower than that in the control group (226.40 6 15.65 g, P < 0.01). There was no significant difference of mean body weight between group I and group II. At 18 wk, the mean body weight was 183.40 6 15.39 g in group III, which was lower than that in the control group (234.80 6 14.13 g, P < 0.01). There were no significant differences in the consumption of DEN between the control groups and the FR-treated groups. DISCUSSION
Curative treatments currently available for HCC include surgical resection, liver transplantation, and percutaneous ablation. However, many patients have TABLE 1 Comparison of the Mean Maximum Diameter and the Mean Number of Red Nodules Red nodules Maximal diameter (mm) Control group (12 wk) Group I Group II Control group (18 wk) Group III a
P < 0.01. P < 0.01.
b
9.2 6 1.30a – 4.4 6 1.14a 18.2 6 2.17b 10.6 6 2.59b
Number 9.6 6 1.14a – 2.6 6 0.09a 15.2 6 2.17b 7.2 6 1.48b
ISHII ET AL.: ANTI-ANGIOGENIC THERAPY ON HCC
73
FIG. 3. Serum AFP levels in the control and FR-treated groups: (A) The serum AFP level of the control group rose at 6 wk and increased further up to 18 wk. The serum AFP levels of group I or II were significantly lower at 12 wk than that of the control group. (B) The serum AFP level of group III was significantly lower than that of the control group. a, b, cP <0.01, dP <0.01. Cont.: Control, W: weeks.
advanced HCC for which these treatments are not indicated. These patients are candidates for palliative treatment, which includes transarterial embolization, hormone therapy, interferon therapy, and chemotherapy. The response rate achieved with systemic chemotherapy is generally less than 30% [12], and patients often tolerate such a treatment poorly. Accordingly, we need new treatment strategies for HCC. In 1971, Folkman suggested the use of anti-angiogenic agents to inhibit tumors, a proposal which led to a new modality of anticancer treatment [13]. HCC is a typical hypervascular tumor; many angiogenic factors have been studied in this cancer, and several anti-angiogenic agents have been tested in animals and patients [14–16]. Angiogenesis is regulated by a balance between stimulators and inhibitors under normal conditions [17]. Some regulators of angiogenesis specifically target ECs, such as vascular endothelial cell growth factor (VEGF), angiopoietins, and endostatin, while others have multiple biological functions in addition to their role in angiogenesis, such as various chemokines that are also involved in the inflammatory response [14]. The relation between inflammation and cancer has attracted increasing attention because of the observation that chronic inflammation is usually associated with cancer [18]. A model of HCC induced by DEN was used in this study because it resembles human HCC [2]. A review of anti-angiogenic therapy for HCC showed that examination of anti-angiogenic treatment during hepatocarcinogenesis has not been performed before [14]. Therefore, this was taken into consideration in our study. Three-dimensional changes of the vasculature during hepatocarcinogenesis induced by DEN were studied
by SEM observation of vascular casts created by infusing red resin via the hepatic artery and blue resin via the portal vein. Nodules that appeared at 6 wk were filled with blue resin similar to the surrounding sinusoids, and the vascular casts of these nodules were roundish and irregular. Thus, changes of the vascular
FIG. 4. Serum IL-8 levels of the control and FR-treated groups: IL-8 was already increased at 6 wk, and rose further up to 18 wk. The serum IL-8 levels of groups I or II were significantly lower at 12 weeks than that of the control group. The serum IL-8 level of the group III was significantly lower than that of the control group. On the other hand, there was no significant difference between group III and the control 12 wk. There was a significant difference of serum IL-8 between group I and the control group 6 weeks. On the other hand, there was no significant difference between group I and normal animals. a, b, cP < 0.01, d, fP < 0.05, eP ¼ 0.252, gP ¼ 0.078. Cont.: Control, W: weeks.
74
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
FIG. 5. Scanning electron microscopy of vascular casts: (A) Normal liver. HA: hepatic arteriole, PV: portal venule. (B) High power view of a normal sinusoid. CHV: central hepatic vein. (C) Hyperplastic nodules at 6 wk. (D) Blood sinusoids in the control group at 12 wk. (E) High power view of blood sinusoids in the control group at 12 wk. (F) Blood sinusoids in the control group at 18 wk. (G) Angiogenesis at the border of a red nodule. (H) Sinusoidal architecture in group III. (I) Sinusoidal architecture in group I.
architecture were already observed at this time. The red nodules seen at 12 wk had incompletely anastomosing sinusoids with a coral-like appearance. These findings are typical of so-called ‘‘blood sinuses’’ and the red nodules were shown to be HCCs histopathologically. Furthermore, angiogenesis was prominent at the edges of these nodules. Red nodules (HCCs) at 18 wk exclusively contained arterial red resin that outlined irregular, mesh-like, anastomosing sinusoids. The importance of angiogenesis in the process of hepatocarcinogenesis was clarified by this morphological study. Endothelial lining cells usually express CD31 and CD34, but these markers are rarely associated with normal sinusoidal ECs [16]. The so-called capillarization of tumor-associated vessels may either result from the development of new vessels through the process of tumor angiogenesis or from the alteration of pre-existing sinusoids in response to tumor-induced
changes of the microenvironment [19, 20]. We immunohistochemically examined the expression of CD31, which is an endothelial-specific marker. CD31 binds with PECAM-1, a cell–-cell adhesion molecule of the immunoglobulin superfamily expressed by most endothelial cells [21]. In the control group, MVD was increased from 6 wk, and increased further by 18 wk. Furthermore, IL-8 (an angiogenic factor) was elevated at 6 wk and increased further after that with the development of the HCCs. Our findings strongly suggest that angiogenesis is closely related to hepatocarcinogenesis and tumor growth. Because endothelial cells play a central role in angiogenesis, FR (which directly acts on endothelial cells) was selected for this study. FR is an angiogenesis inhibitor that has been confirmed to suppress neovascularization through its inhibitory effect on EC proliferation both in vitro and in vivo [3, 4]. FR was synthesized by
ISHII ET AL.: ANTI-ANGIOGENIC THERAPY ON HCC
75
FIG. 6. The assessment of microvessel density (MVD) in tumor tissues. MVD was stained by CD31 immunohistochemistry using diaminobenzidine (DAB) and the image was displayed on the monitor with a Zeiss Axiovision 2.05. (A) Control 12 wk (3200), (B) group II (3200).
modification of the fermentation products of the fungus Scolecobasidum arenarium at Fujisawa Pharmaceutical Co., Ltd. (Tsukuba, Japan) [3]. In addition to its effect on HCC, the optimum timing of FR administration was examined during the process of hepatocarcinogenesis and tumor growth. No HCCs developed in group I, suggesting the usefulness of FR as a chemopreventive agent. As data from group II showed, even if anti-angiogenic treatment is started before the development of HCC, it is impossible to control the tumor subsequently. The point of debate about chemoprevention with anti-angiogenic therapy
FIG. 7. Immunohistochemical analysis of microvessel density (MVD). FR-118487 significantly reduced the MVD. Compared with the control group at 12 wk, the MVD was reduced by 35% in group II and by 49% in group I. There was also a significant difference of MVD between group I and group II. Compared with the control group, MVD in group III was reduced by 40% at 18 wk. There was no significant difference of MVD between group II and the control group at 18 weeks. a, b, dP < 0.01, cP < 0.05, eP ¼ 0.0143.
is when to start such treatment for patients with chronic liver disease before the development of HCC. However, the results for groups I and II indicate that starting anti-angiogenic treatment as soon as possible is better for chemoprevention. The mean maximum size and number of red nodules in groups II and III were reduced compared with the control group. On the other hand, there were no significant differences in the maximum size and number of red nodules between the control group at 12 wk and group III at 18 wk. This suggests that FR was able to induce a dormant state of HCC. Serum AFP levels also reflected these effects of FR. Examination of hepatic vascular casts and comparison with the control group showed that there was less irregularity blood sinuses in groups II and III and avascular areas were conspicuous. Angiogenesis to create feeding vessels for the red nodules was also inhibited. In small hyperplastic nodules from group I, the sinusoidal architecture was closer to that of normal liver than in the other groups. Immunohistochemistry for CD31 showed that MVD was significantly lower in groups II and III than in the respective control groups. Measurement of IL-8 also supported the results of immunohistochemical examination. It has been reported that IL-8 induces chemotactic activity on ECs and active angiogenesis in vivo [22]. Potently angiogenic, IL-8 is constitutively produced by various tumor cell lines, including HCC lines [23]. Biological markers are needed to demonstrate the mechanisms of anti-angiogenic therapy and follow up its effects because conventional criteria for complete response (CR) or partial response (PR) fail to show the effects of such treatment. Such markers may include the serum concentrations of angiogenic factors. In this study, the serum level of IL-8 increased before the development of HCCs and rose further after the tumors appeared. In the FR-treated groups, IL-8 levels were lower compared with those in the control group. Our investigation of IL8 in HCC patients showed that it was elevated with
76
JOURNAL OF SURGICAL RESEARCH: VOL. 158, NO. 1, JANUARY 2010
tumor development, and the elevation was considered to be important for hepatocarcinogenesis [24]. Accordingly, IL-8 may possibly serve as a marker for the effect of anti-angiogenic treatment on HCC (including chemoprevention). However, further studies are needed to establish anti-angiogenic therapy for chemoprevention using IL-8 as an index. The inhibition of weight gain is a major adverse effect of this therapy. The optimum dose, duration, and route of administration for clinical application still need to be determined. In regard to this, the continuous infusion of FR suppressed liner metastasis without any adverse effects [25]. FR-treated rats did not show side effects such as hair loss or infection that occur with conventional chemotherapy. In conclusion, FR markedly inhibited the development and progression of HCC through inhibition of angiogenesis in an animal model of HCC induced by DEN. Anti-angiogenic therapy may be effective for chemoprevention and tumor dormancy in patients who have HCC associated with chronic liver disease. Furthermore, IL-8 may be a useful marker for monitoring antiangiogenic treatment in HCC patients. ACKNOWLEDGMENTS The authors gratefully acknowledge Professor H. Hano, Department of Pathology, Jikei University School of Medicine, for excellent technical support.
REFERENCES 1. Li CY, Shan S, Huang Q, et al. Initial stages of tumor cellinduced angiogenesis: Evaluation via skin window chambers in rodent models. J Natl Cancer Inst 2000;92:143. 2. Masaka A. Sequential immunohistochemical study of early hepatocellular carcinoma in rat liver during carcinogenesis induced by diethylnitrosamine. St Mariana Med J (Japanese) 1992;20:472. 3. Otsuka T, Okawa T, Shibata T, et al. A new potent angiogenesis inhibitor, FR-118487. J Microbiol Biotechnol 1991;1:163. 4. Terano H, Shibata T, Otsuka T. Angiogenesis inhibitors of microbial origin. Drug Future 1993;18:239. 5. Skinner SA, Tutton PJM, O’Brien PE. Microvascular architecture of experimental colon tumors in the rat. Cancer Res 1990; 50:2411. 6. Tsuda H, Masui T, Inoue T. Three-dimensional observation of vasculature and blood supply during hepatocarcinogenesis in rats. Saishinigaku (Japanese) 1986;41:67. 7. Guesdon JL, Ternynck T, Avrames S. The use of avidin-biotin interaction in immunoenzymatic techniques. J Histochem Cytochem 1979;27:1131.
8. Wamke R, Levy R. Detection of T and B cell antigens with hybridoma monoclonal antibodies: A biotin-avidin-horseradish peroxidase method. J Histochem Cytochem 1980;28:771. 9. Yoshiji H, Kuriyama S, Ways DK, et al. Protein kinase C lies on the signaling pathway for vascular endothelial growth factormediated tumor development and angiogenesis. Cancer Res 1999;59:4413. 10. Kimura H, Nakajima T, Karawa K, et al. Angiogenesis in hepatocellular carcinoma as evaluated by CD34 immunohistochemistry. Liver 1998;18:14. 11. Moriguchi M, Nakajima T, Kimura H, et al. The copper chelator trientine has an anti-angiogenic effect against hepatocellular carcinoma, possibly through inhibition of interleukin-8 production. Int J Cancer 2002;102:445. 12. Thomas MB, Chadha R, Glover K, et al. Phase 2 study of erlotinib with unresectable hepatocellular carcinoma. Cancer 2007; 110:1059. 13. Folkman J. Tumor angiogenesis: Therapeutic implications. N Engl J Med 1971;285:1182. 14. Hui CS, Zaho YT. Angiogenesis in hepatocellular carcinoma: The retrospectives and perspectives. J cancer Res Clin Oncol 2004;130:307. 15. Zhong LZ, Zhi SL, Quan S. Effects of thalidomide on angiogenesis and tumor growth and metastasis of human hepatocellular carcinoma in nude mice. World J Gastroenterol 2005;11:216. 16. Schwartz M, Roayaie S, Knostadoulakis M. Strategies for the management of hepatocellular carcinoma. Nat Clin Pract Oncol 2007;4:424. 17. Glaser BM. Extracellular modulating factors and the control of intraocular neovascularization: An overview. Arch Opthamol 1988;106:603. 18. Lisa MC, Werb Z. Inflammation and cancer. Nature 2002; 420:860. 19. Nakamura S, Muro H, Suzuki S, et al. Immunohistochemical studies on endothelial cell phenotype in hepatocellular carcinoma. Hepatology 1997;26:407. 20. Frachon S, Gouysse G, Dumortier J, et al. Endothelial cell marker expression in dysplastic lesions of the liver: An immunohistochemical study. J Hepatol 2001;34:850. 21. Newman P, Berndt M, Gorski J, et al. PECAM-1 (CD31) cloning and relation to adhesion molecules of the immunoglobulin gene superfamily. Science 1990;247:1219. 22. Arenberg DA, Kunkel SL, Polverini PJ, et al. Inhibition of interleukin-8 reduces tumorigenesis of human non-small cell lung cancer in SCID mice. J Clin Invest 1996;97:2792. 23. Kubo F, Ueno S, Hiwatashi K, et al. Interleukin 8 in human hepatocellular carcinoma correlates with cancer cell invasion of vessels but not with tumor angiogenesis. Ann Surg Oncol 2005;12:800. 24. Ishii Y, Nakasato Y, Yanagisawa S, et al. Dose anti-angiogenic treatment make any contribution in cases of primary hepatocellular carcinoma? With special reference to chemoprevention, tumor dormancy and tumor shrinkage. Bioterapy (Japanese) 2002;16:396. 25. Ueoka M. Inhibitory effects of the angiogenesis inhibitor FR118487 on liver metastasis in a rabbit colon cancer model by continuous intraarterial infusion. Jpn J Gastroenterol Surg (Japanese) 1999;32:983.