O

E

YD

a

ARRA

KASSCC

I

w(asair

iafstaffphdtq

0d

Pathology – Research and Practice 206 (2010) 725–730

Contents lists available at ScienceDirect

Pathology – Research and Practice

journa l homepage: www.e lsev ier .de /prp

riginal Article

ndoglin (CD105) expression and angiogenesis status in small cell lung cancer

ukari Takase, Keita Kai, Masanori Masuda, Michiaki Akashi, Osamu Tokunaga ∗

epartment of Pathology and Biodefense, Saga University, Faculty of Medicine, Nabeshima 5-1-1, Saga City, Saga 849-8501, Japan

r t i c l e i n f o

rticle history:eceived 25 November 2009eceived in revised form 24 April 2010ccepted 24 May 2010

eywords:ngiogenesis

a b s t r a c t

It is well established that angiogenesis is crucial for tumor development and progression. Among theangiogenesis immunomarkers defined to date, endoglin (CD105) has been shown to be a useful markerof angiogenesis. To investigate the degree of angiogenesis status in small cell lung cancer (SCLC) tissue,we assessed 35 cases of SCLC at autopsy using immunohistochemical staining of CD31 and CD105. Theintratumoral area, peritumoral area, and background pulmonary alveoli were then observed under lowmagnification, and the microvessel density (MVD) for each area was determined. The MVD-CD31 was the

mall cell lung cancerCLCD105D31

highest in the background alveoli, followed by the intratumoral and peritumoral areas. The MVD-CD105was highest in the intratumoral area, followed by the peritumoral area and the background lung. The ratioof CD105/CD31 revealed that almost 78% of the intratumoral area, 63% of the peritumoral area, and 4.6% ofthe background lung alveoli were newly formed and expressed CD105. This result indicated that SCLC ispredominantly supported by newly formed vessels that are generated by CD105-mediated angiogenesis.These findings suggest that anti-angiogenic therapy, especially CD105-targeting, may prove an effective

form of SCLC treatment.

ntroduction

Lung cancer is one of the leading causes of cancer deaths world-ide. Most of the lung cancers are non-small cell lung cancer

NSCLC) and small cell lung cancer (SCLC), representing 13–20% ofll lung cancers [23], and are considered the most chemo-sensitiveolid tumors. Despite this, the vast majority of patients eventu-lly experience a relapse, and, as a result, the median survival times only 10–14 months [13,27]. Unfortunately, there is no effectiveadical therapy for SCLC available at present.

Angiogenesis, the formation of new blood vessels from preex-sting vessels [5] or circulating endothelial progenitor cells [1], ismultistep process that promotes metastasis involving the spread

rom the primary tumor through blood and lymphatic vessels andubsequent growth of the secondary tumor at the anchored dis-ant organs. Blocking tumor angiogenesis has been proven to ben important strategy for cancer therapy, and has been approvedor clinical use. Such agents have been shown to be beneficialor advanced cancer patients [9,15]. Antibodies directed againstanendothelial cells, such as anti-CD31 and anti-CD34 antibodies,

ave also been used in the evaluation of angiogenesis. These panen-othelial antibodies react with both the newly forming vessels andhe normal vessels trapped within the tumor tissues. Therefore, theuestion arises as to whether these panendothelial markers are

∗ Corresponding author. Tel.: +81 952 34 2230; fax: +81 952 34 2055.E-mail address: tokunao@cc.saga-u.ac.jp (O. Tokunaga).

344-0338/$ – see front matter © 2010 Elsevier GmbH. All rights reserved.oi:10.1016/j.prp.2010.05.015

© 2010 Elsevier GmbH. All rights reserved.

ideal for the evaluation of tumor angiogenesis. Endoglin (CD105)has proven to be a useful marker of angiogenesis. CD105 servesas a receptor for transforming growth factor (TGF)-�1 and TGF-�3, and modulates TGF-� signaling through its interaction withTGF-� receptor I and TGF-� receptor II (TGF-�RII) [2]. The TGF-�/ALK signaling pathway is activated by TGF-�1, TGF-�RII, andCD105 and promotes the proliferation of the endothelium [8].CD105 is highly expressed on proliferating endothelial cells of boththe peri- and intratumoral blood vessels that participate in tumorangiogenesis, and is either weakly or negatively expressed in thevascular endothelium of normal tissues [3,17]. Several studies haveindicated the usefulness of CD105 as a target for anti-angiogenictherapy [6,16] and its superiority in prognosis compared to otherpanendothelial markers such as CD31 or CD34 for various types oftumors [4,14,22,24,30], including NSCLC [26]. However, to the bestof our knowledge, no previous studies have investigated the degreeof CD105-dependent angiogenesis status in SCLC tissue.

The aim of this study was therefore to assess the CD105-dependent angiogenesis status and the potential of CD105anti-angiogenic therapy in SCLC.

Materials and methods

Patient samples and tissue collection

Thirty-five SCLC samples collected at autopsy at Saga UniversityHospital between 1982 and 2007 were evaluated in the currentstudy. All 35 samples were positive for at least one of the fol-

7 arch a

l(btsrcmtU

I

tsftpCmtBsCr(

E

tbtatuamfma[TM

E

wivp

D

vidfpBos

26 Y. Takase et al. / Pathology – Rese

owing markers: CD56, chromogranin A, neuron-specific enolaseNSE), and synaptophysin. Collected samples were fixed in 10%uffered formalin and processed to paraffin. Two blocks of lungissue containing SCLC and invasion-free background lung wereelected for each patient, and 4-�m sections were collected foroutine hematoxylin and eosin (H&E) staining and immunohisto-hemistry. All clinical information was extracted from the patient’sedical records. This study was carried out under the guidelines of

he Ethics Committee for Human Study, Faculty of Medicine, Saganiversity.

mmunohistochemistry

The immunohistochemical analysis was performed followinghe previously reported method [17]. For antigen retrieval, thelides were submerged in EDTA (pH 8) and heated in a microwaveor 6 min (for CD105, TGF-�1, and TGF-�RII) or incubated with pro-einase K for 10 min (for CD31) at room temperature. The followingrimary antibodies were used: mouse monoclonal anti-humanD105 (1:20, clone 4G11; Novocastra, Newcastle, UK), mouseonoclonal anti-human CD31 (1:30, clone JC70A; Dako Cytoma-

ion, Glostrup, Denmark), TGF-�1 (1:100, clone sc-146; Santa Cruziotechnology, Santa Cruz, CA, USA), and TGF-�RII (1:200, clonec-400; Santa Cruz Biotechnology). The Envision+® System (Dakoytomation) was used as the second antibody, and the antibodyeaction was visualized with diaminobenzidine tetrahydrochloride0.02%). Cell nuclei were counterstained with hematoxylin.

valuation of the CD31- and CD105-positive vessels

Three immunostained vessel hotspots [19] were identified athe intratumoral area, the peritumoral area and the invasion-freeackground pulmonary alveoli, and examined at low magnifica-ion. The total number of positive vessels located in the peritumoralnd intratumoral areas was assessed under 200× magnification bywo pathologists, while that in the background alveoli was assessednder 400× magnification because the alveolar lesion has too manynd too small vessels to be counted at 200× magnification. Theicrovessel density (MVD) was calculated as the average count

rom the three hotspot fields of view. Because CD105 is a specificarker of newly formed and activated small blood vessels [7,17]

nd has emerged as one of the targets for anti-angiogenesis therapy7], CD105, as MVD-CD105, was used for analysis of angiogenesis.o evaluate the extent of angiogenesis, the ratio of MVD-CD105 toVD-CD31 was calculated.

valuation of the TGF-ˇ1 and TGF-ˇRII immunohistochemistry

The immunohistochemical evaluation of TGF-�1 and TGF-�RIIas performed independently by three pathologists. The stain

ntensity was evaluated in tumor cells and endothelium of tumoressels in each case and categorized as negative, faintly positive,ositive, or strongly positive.

ouble immunofluorescence staining

Double immunofluorescence staining was performed as pre-iously reported [17] to confirm the results of conventionalmmunohistochemical staining. Briefly, the tissue slides wereeparaffinized and soaked in 0.01 M citrate buffer (pH 6.0) at 90 ◦C

or 40 min for antigen retrieval. Slides were incubated first with therimary rabbit polyclonal antibody against human CD105 (Springerioscience, Pleasanton, CA, USA), followed by FITC-labeled sec-ndary goat anti-rabbit IgG (Genetex Inc., Irvin, CA, USA). Theecond immunoreaction was carried out with primary monoclonal

nd Practice 206 (2010) 725–730

anti-human CD31, followed by Cy3-labeled secondary sheep anti-mouse IgG antibody (Chemicon International, Temecula, CA, USA).FITC- and Cy3-labeled samples were examined using a confocalLaser Scanning Microscope LSM5 Pascal (Carl Zeiss Microimaging,Jena, Germany). To detect nonspecific-antibody binding, controlsections were incubated with either normal murine or rabbitserum, or phosphate-buffered saline instead of primary antibody.No staining was observed in these control samples.

Statistical analysis

Comparisons of MVD and the ratio of MVD-CD105 to MVD-CD31were analyzed using Student’s t-test. Statistical analysis was under-taken using JMP software program version 8 (SAS Institute, Cary,NC, USA). Values of p < 0.05 were considered statistically significant.

Results

Clinicopathological features

The 35 cases evaluated in this study included 29 male and 6female patients. The median age at death was 69 years (range38–87), and all patients who entered the study had died of SCLC.Eight of the cases were pathologically diagnosed as SCLC by cytol-ogy, 25 by biopsy, and two at autopsy. The median survival time ofthe pre-diagnosed cases was 5.8 months (range 0.33–34.4). Elevenpatients underwent chemotherapy, 1 underwent radiotherapy, 17underwent chemo-radiotherapy, and 6 did not receive any therapy.

Expression of CD31 and CD105 in the background lung alveoli andperitumoral and intratumoral areas

CD105 and CD31 were expressed in the endothelium of bloodvessels, but not in SCLC tumor cells. In the background lung alve-olus, CD31 was widely expressed in the endothelium of bloodvessels; CD105 was rarely expressed in this area (Fig. 1a–c). Thenumber of CD105-positive vessels was increased at the tumorfront (Fig. 1d–f), and almost all intratumoral blood vessels wereCD105-positive. This pattern of expression was confirmed inthe double-fluorescence immunohistochemical staining analysis(Fig. 1g–i). In the intratumoral area, CD105 was co-expressed withCD31 in blood vessels (Fig. 2a–f). This expression pattern was alsoconfirmed by double-fluorescence immunohistochemical staining(Fig. 2g–i).

Comparison of MVD-CD31 and MVD-CD105 in the backgroundlung alveoli, peritumoral area, and intratumoral area

The mean MVD-CD31 in the background lung alveolus was63.0 ± 29.4, while that in the tumor front and intratumoral area was23.7 ± 13.6 and 35.8 ± 18.0, respectively (Fig. 3). Significant differ-ences were observed between the peritumoral and intratumoralareas (p = 0.02). The MVD of background alveoli was counted in thefield of 400× magnification; therefore, it is statistically incompara-ble to the MVD of the peritumoral and intratumoral areas, which iscounted in the field of 200× magnification.

The mean MVD-CD105 in the background lung alveolus was2.2 ± 4.2, while that of the peritumoral and intratumoral areas

was 12.3 ± 5.2 and 26.2 ± 14.7, respectively (Fig. 4). Significant dif-ferences were observed between the background alveoli and theperitumoral area (p < 0.0001), between the background alveoli andthe intratumoral area (p < 0.0001), and between the peritumoraland intratumoral areas (p < 0.0001).

Y. Takase et al. / Pathology – Research and Practice 206 (2010) 725–730 727

Fig. 1. (a–c) Representative images of the background pulmonary alveoli. (a) H&E staining (magnification 400×). (b) Immunohistochemical staining of CD31 indicating thenumerous CD31-positive vessels (400×). (c) Immunohistochemical staining of CD105 demonstrating almost all vessels that are negative for CD105 (400×). (d–f) Representativeimages of the peritumoral area. (d) H&E staining (400×). (e) Immunohistochemical staining of CD31 (400×). (f) Immunohistochemical staining of CD105 demonstratingp imm1 ut onr the a

T

waiaap

I

(pniei(4s

i

ositive vessels that are sporadically distributed (400×). (g–i) Double-fluorescence00×). Both the pulmonary alveoli and tumor vessels are CD31-positive (h; 100×), beferences to color in this figure legend, the reader is referred to the web version of

he ratio of MVD-CD105 to MVD-CD31

The ratio of MVD-CD105 to MVD-CD31 in the background alveolias 0.046 ± 0.084, while that of the peritumoral and intratumoral

reas was 0.63 ± 0.32 and 0.78 ± 0.33, respectively (Fig. 5). Signif-cant differences were observed between the background alveolind the peritumoral area (p < 0.0001), between the backgroundlveoli and the intratumoral area (p < 0.0001), and between theeritumoral and the intratumoral areas (p = 0.0197).

mmunohistochemistry of TGF-ˇ1 and TGF-ˇRII

Tumor cells were evaluated to be TGF-�1-negative in 26 cases74.3%), faintly positive in 8 (22.9%), positive in 0, and stronglyositive in 1 (2.9%). The endothelium was evaluated to be TGF-�1-egative in 10 cases (28.6%), faintly positive in 13 (54.3%), positive

n 5 (14.3%), and strongly positive in 1 (2.9%). Tumor cells werevaluated to be TGF-�RII-negative in 11 cases (31.4%), faintly pos-tive in 13 (37.1%), positive in 4 (11.4%), and strongly positive in 7

20%). The endothelium was evaluated to be TGF-�RII-negative incases (11.4%), faintly positive in 21 (60%), positive in 3 (8.6%), and

trongly positive in 7 (20%) (Fig. 6).Fig. 7 shows the representative images of immunohistochem-

stry of TGF-�1 and TGF-�RII.

unohistochemical staining of CD31 (red) and CD105 (green), and H&E staining (g;ly tumor vessels express CD105 (i; merged image, 100×). (For interpretation of therticle.)

Discussion

SCLC is generally treated with a combination of chemotherapyand radiotherapy [23]. SCLC tissue samples are not readily availablefor research, making pathology-based research of SCLC difficult.To overcome this obstacle, the present study involved the use ofsamples collected at autopsy. Although the majority of our casesunderwent clinical treatment, the median survival time was only5.8 months. At present, effective radical therapy for SCLC is unavail-able, and thus, at present, novel therapeutic approaches are beinginvestigated to improve patient prognosis.

Blocking tumor angiogenesis has been proven to be an impor-tant strategy for cancer therapy, and anti-angiogenesis cancertherapies, i.e., with the VEGF inhibitor bevacizumab (Avastin,Genentech, San Francisco, CA, USA) as the first generation drug [19],and the tyrosine kinase inhibitors sorafenib (Nexavar, Bayer, Lev-erkusen, Germany) [29] and sunitinib malate (Sutent, Pfizer, NewYork, NY, USA) [20] as the second generation drugs, have beendeveloped and approved for clinical use. These agents have beenshown to be beneficial for advanced cancer patients [17]. In this

study, we focused on the possibility of anti-angiogenic therapy inSCLC. When bevacizumab treatment is combined with paclitaxeland carboplatin in patients with advanced NSCLC, an improvementin overall survival is often observed [10,12]. However, the applica-tion of bevacizumab for SCLC has been poorly investigated.

728 Y. Takase et al. / Pathology – Research and Practice 206 (2010) 725–730

Fig. 2. (a–i) Representative images of the intratumoral area. (a) H&E staining (magnification 40×). (b) Immunohistochemical staining of CD31 (40×). (c) Immunohistochemicals g of CDs ratum( he eno sion o

cwTtsCmi

Fto

may be an ideal target for anti-angiogenic therapy. Many preclinical

taining of CD105 (40×). (d) H&E staining (200×). (e) Immunohistochemical stainintaining of CD105 demonstrating CD105-positive vessels that are dominant in the inth; red) and CD105 (i; green), and the merged image (g). In the intratumoral area, tf the references to color in this figure legend, the reader is referred to the web ver

We reported previously that CD105 is highly expressed inancerous lesions located in various organs, including the lung,hereas lower levels of VEGF expression were observed [18].

hese findings suggest that CD105 may prove more effective as aherapeutic target than VEGF. Additional studies have also demon-

trated that the systemic administration of naked anti-humanD105 monoclonal antibody suppresses established tumors in ani-al models, and that its efficacy is markedly enhanced when used

n combination with a chemotherapeutic drug administered in

ig. 3. MVD-CD31 is the highest in the background lung alveoli, followed byhe intratumoral area and then the peritumoral area. Significant differences arebserved between the peritumoral and the intratumoral areas (*p = 0.02).

31 demonstrating numerous slit-shaped vessels (200×). (f) Immunohistochemicaloral area (200×). (g–i) Double-fluorescence immunohistochemical staining of CD31

dothelium is stained positive for both CD31 and CD105 (200×). (For interpretationf the article.)

an anti-angiogenic drug dose schedule [3,25]. Furthermore, anti-CD105 antibody was found to not only inhibit the progression ofa colonic cancer xenograft, but also hematogenous metastasis inSCID mice [16]. In combination, these results suggest that CD105

studies using various clones of anti-CD105 monoclonal antibodieswith animal models of cancer cell lines have been reported [8,28].At present, a phase I, first-in human study with the human/murinechimeric anti-CD105 monoclonal antibody TRC105 is ongoing in

Fig. 4. MVD-CD105 is the highest in the intratumoral area, followed by the peri-tumoral area and then the background alveoli. Significant differences are observedbetween the peritumoral and the intratumoral areas (*p < 0.0001).

Y. Takase et al. / Pathology – Research and Practice 206 (2010) 725–730 729

Fig. 5. Ratio of MVD-CD105 to MVD-CD31 (CD105/CD31). CD105/CD31 is thehighest in the intratumoral area, followed by the peritumoral area and then the back-gaia

pHHdses

wafwbmeootrSv

Fiaw�i

Fig. 7. Representative images of immunohistochemical staining of TGF-�1 andTGF-�RII. (a) Immunohistochemical staining of TGF-�1. Tumor cells are negatively

round alveoli. Significant differences are observed between the background alveolind the peritumoral area (*p < 0.0001), between the background alveoli and thentratumoral area (**p < 0.0001), and between the peritumoral and the intratumoralreas (***p = 0.0197).

atients with refractory advanced or metastatic solid cancer [21].owever, only one report has focused on the role of CD105 in SCLC.ardavella et al. used CD105 as a marker of lymphangiogenesis andemonstrated that CD105 expression and lymphatic invasion areignificantly correlated with patient outcome in SCLC [11]. How-ver, this study failed to refer to the CD105-dependent angiogenesistatus.

In order to evaluate the CD105-dependent angiogenesis status,e performed immunohistochemical analysis of CD31 and CD105

nd found that MVD-CD31 was highest in the background lung,ollowed by the intratumoral area and then the peritumoral area,hile MVD-CD105 was highest in the intratumoral area, followed

y the peritumoral area and then the background lung. As the deter-ination of MVD alone is thought to be insufficient for the precise

valuation of the extent of angiogenesis, we calculated the ratiof MVD-CD105 to MVD-CD31 and found that approximately 78%

f intratumoral and 63% of peritumoral vessels, but only 4.6% ofhe background lung alveoli vessels, were positive for CD105. Thisesult indicates that angiogenesis is crucial for the development ofCLC, and that SCLC is predominantly supported by newly formedessels. Given these findings, we hypothesized that anti-angiogenic

ig. 6. Immunohistochemical analysis of TGF-�1 and TGF-�RII. The stain intensitys evaluated as negative, faintly positive, positive, or strongly positive. The verticalxis is the number of cases. Tumor cells are dominantly stained TGF-�1-negative,hereas the endothelium in the tumor area is stained at various intensities by TGF-1. On the other hand, tumor cells and the endothelium are both stained at various

ntensities by TGF-�RII.

stained by TGF-�1, whereas some areas of endothelium are positively stained byTGF-�1 (arrows). Insert shows the only case of strongly positive staining of TGF-�1 (magnification 200×). (b) Immunohistochemical staining of TGF-�RII. Both thetumor cells and the endothelium (arrow) are positively stained by TGF-�RII (200×).

therapy, especially CD105-targeting, may prove effective for thetreatment of SCLC.

To investigate the angiogenic mechanism in SCLC, we immuno-histochemically evaluated TGF-�1 and TGF-�RII and revealed that74.3% of SCLC were negative for TGF-�1. On the other hand, 72.4% ofthe endothelium samples expressed TGF-�1, 88.6% expressed TGF-�RII, of which 20% were strongly positive. These results indicatethat SCLC usually does not produce TGF-�1, but the endotheliumin SCLC usually expresses TGF-�RII. It is possible therefore thatnon-tumor cells, such as inflammatory cells or fibroblasts, produceTGF-�1. Consequently, CD105 expression in the endothelium mightbe passively enhanced in the tumor microenvironment, which isrich in various cytokines. Thus, the TGF-�/ALK1 signaling pathwaycould be activated by TGF-�RII, CD105, and TGF-�1 to promotethe proliferation of the endothelium. However, our experimentwas insufficient to verify this theory, and further investigation isrequired to determine the precise mechanism of angiogenesis in

SCLC.

In conclusion, this study revealed the cancer-specific expres-sion of CD105 in SCLC. We found that SCLC was nourished mainlyby newly formed vessels that are dependent on CD105-mediatedangiogenesis for their establishment. Therefore, the results of this

7 arch a

sa

A

i

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

tyrosine kinases involved in tumor progression and angiogenesis, Cancer Res.64 (2004) 7099–7109.

30 Y. Takase et al. / Pathology – Rese

tudy indicate the therapeutic potential of CD105-targeted anti-ngiogenic treatment in patients with SCLC.

cknowledgement

We would like to thank Mr. F. Mutoh for his contribution to themmunohistochemical studies.

eferences

[1] T. Asahara, H. Masuda, T. Takahashi, C. Kalka, C. Pastore, M. Silver, M. Kearme,M. Magner, J.M. Isner, Bone marrow origin of endothelial progenitor cellsresponsible for postnatal vasculogenesis in physiological and pathological neo-vascularization, Circ. Res. 85 (1999) 221–228.

[2] N.P. Barbara, J.L. Wrana, M. Letarte, Endoglin is an accessory protein thatinteracts with the signaling receptor complex of multiple members of thetransforming growth factor-superfamily, J. Biol. Chem. 274 (1999) 584–594.

[3] S.E. Duff, C. Li, J.M. Garland, S. Kumar, CD105 is important for angiogenesis:evidence and potential applications, FASEB J. 17 (2003) 984–992.

[4] O. Erdem, C. Taskiran, M.A. Onan, M. Erdem, H. Guner, O. Ataoglu, CD105expression is an independent predictor of survival in patients with endometrialcancer, Gynecol. Oncol. 103 (2006) 1007–1011.

[5] J. Folkman, Angiogenesis in cancer, vascular, rheumatoid, and other disease,Nat. Med. 1 (1995) 149–153.

[6] E. Fonsatti, M. Altomonte, P. Arslan, M. Maio, Endoglin (CD105): a target foranti-angiogenetic cancer therapy, Curr. Drug Targets 4 (2003) 291–296.

[7] E. Fonsatti, M. Maio, Highlights on endoglin (CD105): from basic findingstowards clinical applications in human cancer, J. Transl. Med. 2 (2004) 18.

[8] E. Fonsatti, H.J.M. Nicolay, M. Altomonte, A. Covre, M. Maio, Targeting cancervasculature via endoglin/CD105: a novel antibody-based diagnostic and ther-apeutic strategy in solid tumours, Cardiovasc. Res. 30 (2009) (Epub ahead ofprint).

[9] E. Fonsatti, L. Sigalotti, P. Arslan, M. Altomonte, M. Maio, Emerging role ofendoglin (CD105) as a marker of angiogenesis with clinical potential in humanmalignancies, Curr. Cancer Drug Targets 3 (2003) 427–432.

10] C. Gridelli, P. Maione, A. Rossi, F. De Marinis, The role of bevacizumab in thetreatment of non-small cell lung cancer: current indications and future devel-opments, Oncologist 12 (2007) 1183–1193.

11] G. Hardavella, E. Arkoumani, Y. Dalavanga, P. Galanis, N.J. Agnantis, S. Con-stantopoulos, D. Stefanou, D. Prognostic impact of lymphangiogenesis andlymphatic invasion (CD105 expression) in small cell lung carcinoma, AnticancerRes. 28 (2008) 343–347.

12] L. Horn, A.B. Sandler, Emerging data with antiangiogenic therapies in early andadvanced non-small-cell lung cancer, Clin. Lung Cancer 10 (Suppl. 1) (2009)S7–S16.

13] L.M. Krug, M.G. Kris, K. Rosenzweig, W. Travis, Small cell and other neuroen-docrine tumors of the lung, in: V.T. DeVita, S. Hellman, S.A. Rosenberg (Eds.),

Cancer: Principles & Practice of Oncology, Lippincott Williams and Wilkins,Philadelphia, 2008, pp. 946–971.

14] P.A. Kyzas, N.J. Agnantis, D. Stefanou, Endoglin (CD105) as a prognostic factor inhead and neck squamous cell carcinoma, Virchows Arch. 448 (2006) 768–775.

15] J. Marx, Encouraging results for second generation antiangiogenesis drugs, Sci-ence 308 (2005) 1248–1249.

[

nd Practice 206 (2010) 725–730

16] R. Minhajat, L. Hou, K. Kai, Y. Takase, D. Mori, O. Tokunaga, Anti-CD105 inhibitsprimary cancer growth and secondary hematogenous metastasis in a xenograftmodel, Vasc. Dis. Prev. 6 (2009) 91–96.

17] R. Minhajat, D. Mori, F. Yamasaki, Y. Sugita, T. Satoh, O. Tokunaga, Endoglin(CD105) expression in angiogenesis of colon cancer: analysis using tissuemicroarrays and comparison with other endothelial markers, Virchows Arch.448 (2006) 127–134.

18] R. Minhajat, D. Mori, F. Yamasaki, Y. Sugita, T. Satoh, O. Tokunaga, Organ-specificendoglin (CD105) expression in the angiogenesis of human cancers, Pathol. Int.56 (2006) 717–723.

19] L.G. Presta, H. Chen, S.J. O’Connor, V. Chisholm, Y.G. Meng, L. Krummen, M.Winkler, N. Ferrara, Humanization of an anti-vascular endothelial growth fac-tor monoclonal antibody for the therapy of solid tumors and other disorders,Cancer Res. 57 (1997) 4593–4599.

20] R. Roskoski Jr., Sunitinib: a VEGF and PDGF receptor protein kinase and angio-genesis inhibitor, Biochem. Biophys. Res. Commun. 356 (2007) 323–328.

21] L. Rosen, M.S. Gordon, H.I. Hurwitz, M.K. Wong, B.J. Adams, D. Alvarez, B.K. Seon,B.R. Leigh, C.P. Theuer, Early evidence of tolerability and clinical activity froma phase I study of TRC105 (anti-CD105 antibody) in patients with advancedrefractory cancer, J. Clin. Oncol. 27 (2009) 3518 (Abstract).

22] R.S. Saad, Y. EL-Gohary, E. Memari, Y.L. Liu, J.F. Silverman, Endoglin (CD105)and vascular endothelial growth factor as prognostic markers in esophagealadenocarcinoma, Hum. Pathol. 36 (2005) 955–961.

23] G.R. Simon, A. Turrisi, American College of Chest Physicians, Management ofsmall cell lung cancer: ACCP evidence-based clinical practice guidelines (2ndedition), Chest 132 (Suppl. 3) (2007) 324S–329S.

24] Y. Sugita, Y. Takase, D. Mori, O. Tokunaga, A. Nakashima, M. Shigemori, Endoglin(CD105) is expressed on endothelial cells in the primary central nervous systemlymphomas and correlates with survival, J. Neurooncol. 82 (2007) 249–256.

25] N. Takahashi, A. Haba, F. Matsuno, B.K. Seon, Antiangiogenic therapy of estab-lished tumors in human skin/severe combined immunodeficiency mousechimeras by antiendoglin (CD105) monoclonal antibodies and synergybetween anti-endoglin antibody and cyclophosphamide, Cancer Res. 61 (2001)7846–7854.

26] F. Tanaka, Y. Otake, K. Yanagihara, Y. Kawano, R. Miyahara, M. Li, S. Ishikawa, H.Wada, Correlation between apoptotic index and angiogenesis in non-small celllung cancer: comparison between CD105 and CD34 as a marker of angiogenesis,Lung Cancer 39 (2003) 289–296.

27] N. Thatcher, C. Faivre-Finn, P. Lorigan, Management of small-cell lung cancer,Ann. Oncol. 16 (Suppl. 2) (2005) ii235–ii239.

28] S. Uneda, H. Toi, T. Tsujie, M. Tsujie, N. Harada, H. Tsai, B.K. Seon, Anti-endoglinmonoclonal antibodies are effective for suppressing metastasis and the primarytumors by targeting tumor vasculature, Int. J. Cancer 125 (2009) 1446–1453.

29] S.M. Wilhelm, C. Carter, L. Tang, D. Wilkie, A. McNabola, H. Rong, C. Chen, X.Zhang, P. Vincent, M. McHugh, Y. Cao, J. Shujath, S. Gawlak, D. Eveleigh, B. Row-ley, L. Liu, L. Adnane, M. Lynch, D. Auclair, I. Taylor, R. Gedrich, A. Voznesensky,B. Riedl, L.E. Post, G. Bollag, P.A. Trail, BAY 43-9006 exhibits broad spectrumoral antitumor activity and targets the RAF/MEK/ERK pathway and receptor

30] Y. Yao, T. Kubota, H. Takeuchi, K. Sato, Prognostic significance of microves-sel density determined by an anti-CD105/endoglin monoclonal antibody inastrocytic tumors: comparison with an anti-CD31 monoclonal antibody, Neu-ropathology 25 (2005) 201–206.