Int. J. Cancer: 122, 2805–2810 (2008) ' 2008 Wiley-Liss, Inc.

Expression pattern and circulating levels of endostatin in patients with pancreas cancer € € Daniel Ohlund, Bjarne Ardnor, Mikael Oman, Peter Naredi and Malin Sund* Department of Surgery, Umea˚ University Hospital, SE Ume a, ˚ Sweden Endostatin is a potent inhibitor of angiogenesis that is cleaved from the basement membrane protein type XVIII collagen. Expression of endostatin has recently been shown by Western blot analysis of tissue lysates in normal pancreas and pancreas cancer tissue. We show here that the expression pattern of type XVIII collagen/endostatin is shifted from a general basement membrane staining and is mainly located in the vasculature during tumor progression. This shift in type XVIII collagen/endostatin expression pattern coincides with an up-regulation of MMPs involved in endostatin processing in the tumor microenvironment, such as MMP-3, MMP-9 and MMP-13. The circulating levels of endostatin was analyzed in patients with pancreas cancer and compared to that of healthy controls, as well as after surgical treatment or in a group of nonoperable patients after intraperitoneal fluorouracil (5-FU) chemotherapy. The results show that patients with pancreas cancer have increased circulating levels of endostatin and that these levels are normalized after surgery or intraperitoneal chemotherapy. These findings indicate that endostatin could be used as a biomarker for pancreas cancer progression. ' 2008 Wiley-Liss, Inc. Key words: endostatin; angiogenesis; pancreas cancer; circulation; extracellular matrix

Type XVIII collagen is a heparan sulphate proteoglycan (HSPG) found in most basement membranes (BM).1 Endostatin is a potent antiangiogenic fragment cleaved from the C-terminus of type XVIII collagen by several proteases in the tumor microenvironment such as cathepsin-L, elastase and several matrix metalloproteases (MMPs).2–6. In vitro data has indicated especially MMP-3, -7, -9, -13 and -20 in the processing of endostatin from type XVIII collagen.3 In mice the production of endostatin is initiated by a metal-dependent step resulting in a larger endostatin containing fragment, which then is cleaved further by elastase to yield the functional endostatin molecule, whereas cathepsin-L has been shown to generate endostatin from type XVIII collagen independently of MMP activity.3 Endostatin can also influence the activity of MMPs and matrix remodeling by inhibiting MMP-2 activity.7 Endostatin is found in the circulation at physiological levels of 20–50 ng/ml in serum and 40–100 ng/ml in plasma. The difference in circulating levels between plasma and serum is potentially because of the scavenger property of blood platelets, as endostatin has been shown to be enriched into blood platelets.3,8 Mice deficient of endostatin display increased tumor growth when implanted with cancer cells that do not produce type XVIII collagen, and over-expression of circulating endostatin in transgenic mice leads to reduced tumor growth and vascularization.9 Endostatin is currently in phase II clinical trials to be tested as a potential cancer therapeutic.10,11 Endostatin inhibits tumor growth by several means including inhibition of endothelial cell (EC) proliferation and migration, induction of apoptosis and the G1 arrest of ECs.2,12,13 Many EC surface receptors have been shown to bind endostatin, such as a5b1, avb3, avb5 integrins and glypicans.14–16 The binding of endostatin to a5 integrin leads to the inhibition of the focal adhesion kinase (FAK)/c-Raf/MEK1/2/p38/ERK1 mitogen-activated pathway as well as a disruption of focal adhesions and actin stress fibers by down-regulating RhoA.16–18 The breakdown of the ECM and BM is essential for cancer progression and spreading. MMPs are a family of zinc-dependent Publication of the International Union Against Cancer

proteases that can degrade virtually all extracellular matrix (ECM) and BM components during angiogenesis and carcinogenesis.19 MMPs have also been shown to affect neovascularization directly by regulating EC adhesion, proliferation and migration.19 Both endogenous tissue inhibitors of matrix metalloproteases (TIMPs) and synthetic MMP inhibitors block angiogenesis in vitro and in vivo.20 Although MMPs have traditionally been considered to be proangiogenic and protumorigenic, certain MMPs may also participate in the inhibition of neovascularisation by liberating matrix-derived antiangiogenic substances such as endostatin, tumstatin, arresten and canstatin from their parent molecules in the ECM.21–24 Brammer et al. have recently shown that mature 20 kDa endostatin is found in pancreas cancer tissue by Western blot analysis but not in normal pancreas tissue.25 However, several endostatinrelated peptides of higher molecular weight were present in both tissues. Extracts from normal tissue were able to degrade exogenous endostatin, whereas extracts from cancer could not.25 Interestingly, the addition of elastatinal, a specific inhibitor of elastase, reduced the rate of degradation. The authors conclude that the endostatin is present and stable in pancreatic cancer tissues, but that normal pancreatic tissue expresses enzymes, including elastase, which rapidly degrade endostatin.25 The same authors have subsequently shown that collagen XVIII/endostatin mRNA is expressed in the SUIT-2 pancreatic cell line, and that endostatin is found in the cell-conditioned medium of these cells.26 In a recent publication by Abollahi et al. it has been shown that endostatin treatment of endothelial cells leads to a regulation of multiple genes in a cDNA array that potentially are involved in the angiogenic switch in vitro. Interestingly, the finding was validated in vivo in human pancreas samples including normal, inflamed and cancerous tissue.27 We describe here the expression pattern of various MMPs in the normal and malignantly transformed pancreas in relation to type XVIII collagen/endostatin expression. Our results indicate that several of the MMPs that have been described to be involved in releasing endostatin from type XVIII collagen are up-regulated in pancreas cancer. We also show that in cancer the expression of type XVIII collagen/endostatin is lost from the epithelial BM and endostatin is mainly found in the vascular basement membrane zone. We have also studied the circulating levels of endostatin in patients with pancreas cancer when compared to healthy controls, as well as before and after surgical treatment or in a group of nonoperable patients after intraperitoneal fluorouracil chemotherapy. Our results show that patients with pancreas cancer have increased circulating levels of endostatin and that these levels are normalized after surgery or intraperitoneal chemotherapy. The results indicate that analyzing circulating endostatin levels could be used as a biomarker for pancreas cancer progression. However, further studies are warranted to test this hypothesis. Grant sponsors: Finnish Medical Society Duodecim, Cancerforskningsfonden Norrland, Svenska S€allskapet f€or Medicinsk Forskning SSMF, V€asterbotten County Council and Insamlingsstiftelsen f€or medicinsk forskning at Ume a˚ University. *Correspondence to: Department of Surgery, Ume a˚ University Hospital, SE- 901 85 Ume a, ˚ Sweden. Fax: 146-90-785-1156. E-mail: [email protected] Received 1 August 2007; Accepted after revision 9 January 2008 DOI 10.1002/ijc.23468 Published online 21 March 2008 in Wiley InterScience (www.interscience. wiley.com).

2806

€ OHLUND ET AL.

FIGURE 1 – Expression of type XVIII collagen/endostatin in normal pancreas and pancreas cancer. (a) In normal pancreas type XVIII collagen/ endostatin (in red) is found in most epithelial and vascular basement membranes. Blood vessels are visualized by the endothelial cell marker CD31 (in green) and cell nuclei by DAPI (in blue). (b) In cancer the epithelial basement membrane staining is lost and most endostatin is now found in the tumor VBM zone. Potentially some of the signal is related to the circulating endostatin binding to its EC receptor, as indicated by a colocalization with the endothelial cell marker CD31 (in yellow). The insert in panel B shows a close-up of a blood vessel with colocalization of endostatin and CD31 signal (in yellow). (c, d) Light microscopy images stained with H&E of corresponding areas in normal pancreas and pancreas cancer tissue. (c) In normal pancreas both an islet of Langerhans (marked with L) as well as exocrine pancreas tissue can be seen. (d) The pancreas cancer tissue contains large areas with increased desmoplasia and tumor stroma (marked with S) as well as areas with tumor tissue (marked with T). Magnification 340 in all panels.

donkey-anti-mouse-FITC (1:100) were obtained from Jackson ImmunoResearch. Whole sections were analyzed, and a representative area from the normal pancreas, visualizing both the endocrine and exocrine areas, is shown in the image. From the pancreas cancer samples an area with high angiogenic activity in the tumor is shown.

Material and methods Patients and samples Patients with pancreas cancer were admitted to the Department of Surgery, Ume a˚ University Hospital and evaluated by CT scan, MRI and laparoscopy prior to treatment. Patients who were considered operable underwent a curatively aimed pancreatico-duodenectomy. Patients with a nonresectable or metastatic pancreas cancer were treated with intraperitoneal (IP) chemotherapy with fluorouracil (5-FU).28,29 Plasma samples were collected from patients before (Pre Op; n 5 8) and after surgery (Post Op; >8 weeks, n 5 8), as well as after IP treatment (>12 weeks, n 5 15). Plasma samples were also collected from control patients (n 5 8) who were admitted to the hospital for a nonmalignant disease. The samples were frozen and stored at 280°C until analysis. Pancreas tissue samples from patients with pancreas cancer (n 5 6) and controls (n 5 6) were snap frozen in liquid nitrogen and stored at 280°C until analysis. Informed consent was obtained from all patients. The ethical committee at the Medical faculty of Ume a˚ University approved the study.

Statistics Values are expressed as mean 6 SEM. Statistical analysis was performed by using Student’s t test. The variance of the groups was tested for equality by F-test prior to t test analysis. Two-tailed analysis was used throughout.

Immunohistochemistry Immunohistochemical staining was performed as previously described.9 Six micrometer frozen sections were incubated with the following primary antibodies and dilutions sheep-anti-CD31 (1:100), rabbit-anti-MMP1 (1:100), rabbit-anti-MMP2 (1:100), rabbit-anti-MMP3 (1:100) (all from Chemicon), goat anti-endostatin (1:75), mouse-anti-MMP7 (1:100), mouse-anti-MMP9 (1:100) and mouse-anti-MMP13 (1:100) (all from R&D Systems). Secondary antibodies donkey-anti-sheep-FITC (1:100), donkeyanti-goat-TRITC (1:100), donkey-anti-rabbit-FITC (1:100) and

Results and discussion Altered expression pattern of endostatin in pancreas cancer Type XVIII collagen/endostatin has previously been shown to be expressed in most epithelial BM and vascular basement membranes (VBM) in both developing and mature tissues.1 In the normal pancreas type XVIII collagen/endostatin is found in the basement membrane zones of both the exocrine and endocrine portions of pancreas (Fig. 1a). In the malignantly transformed pancreas the staining pattern of type XVIII collagen/endostatin is shifted so that the BM staining is gradually lost and most signal is found in

ELISA assay Circulating endostatin levels were measured in duplicates from plasma obtained from patients by using the QuantikineTM Human Endostatin Immunoassay (R&D Systems) according to the manufacturer’s protocol.

ENDOSTATIN IN PANCREAS CANCER

2807

FIGURE 2 – Expression pattern of MMPs in normal pancreas and pancreas cancer. (a–c and g–i) In normal pancreas the expression levels of MMPs (in green) are generally low. Almost no staining could be observed for MMP-1 (a), MMP-7 (g), MMP-9 (h) and MMP-13 (i), whereas low to moderate signals could be observed for MMP-3 (c) and MMP-2 (b), respectively. The most intense staining in the normal pancreas was located to the islet of Langerhans of the endocrine pancreas. (d–f and j–l) In pancreas cancer a slight or strong upregulation of signal levels were observed for all MMPs studied. In all cases the increased expression appeared to originate from the cancer cells, although an upregulation of MMP-3 (f) expression also was found in the tumor vasculature as indicated by colocalization with the endostatin staining (in red). It also appears that MMP-3 (f), MMP-9 (k) and MMP-13 (l) expression is more strongly upregulated when compared to MMP-1 (d) and MMP-2 (e). Only weak MMP-7 (j) expression could be observed in the pancreas cancer samples. Cell nuclei are visualized by DAPI (in blue) throughout. Magnification 340 in all panels.

the VBM zones of the tumor (Fig. 1b). Most likely the signal in the VBM zone is partly because of staining of type XVIII collagen in the VBM of the tumor vasculature, as well as by the liberated circulating endostatin molecules binding to the EC receptor as indicated by the overlapping signal of endostatin and EC marker CD31 in the tumor vasculature (Fig. 1b). It has previously been shown that circulating endostatin binds to a5b1 integrin on the

EC surface leading to the activation of antiproliferative and proapoptotic signal and thus an antiangiogenic effect.16 Upregulation of MMP expression colocalises with the shift of endostatin staining in pancreas cancer Several MMPs, such as MMP-3, -7, -9, -13 and -20 have been shown in vitro to be able to process endostatin from the type

€ OHLUND ET AL.

2808

TABLE I – PATIENT CHARACTERISTICS

Sex (% females to males) Primary diagnose (%) Average age (years) p-value versus control p-value surgery versus i.p.chemotherapy Average weight (kg) p-value versus control p-value surgery versus i.p. chemotherapy

Control

Surgery

I.p chemotherapy

62/38 Gallstone disease (75%) Diverticular disease (25%) 50.8 – –

50/50 Pancreas adenocarcinoma (87.5%) Papillarycarcinoma (12.5%) 63.8 0.06 –

54/46 Pancreas adenocarcinoma (100%)

72.1 – –

64.5 0.29 –

67.9 0.42 0.48

60.5 0.09 0.43

I.p., intra peritoneal.

XVIII collagen parent molecule.3 We have analyzed the expression pattern of MMP-1, -2, -3, -7, -9 and -13 in normal pancreas and pancreas cancer in order to see whether an up-regulation of expression can be observed to coincide with the described change in the endostatin expression pattern in cancer. In normal pancreas low levels of MMP-2 and MMP-3 staining could be observed and most of this signal was located to the islets of Langerhans (Figs. 2b and 2c). No staining could be observed for MMP-1, MMP-7, MMP-9 and MMP-13 in the normal pancreas (Figs. 2a, 2g–2i). An upregulation of the expression of all analyzed MMPs in pancreas cancer was observed (Figs. 2d–2f and 2j–2l) as has previously also been shown by others.30,31 For all MMPs analyzed it appears that the signal is located mainly to the cancer cells. Interestingly, a strong upregulation in expression of MMP-3, MMP-9 and MMP13 was observed in pancreas cancer, as these MMPs have all been found to effectively liberate endostatin from type XVIII collagen.3 The increase in MMP-3 signal was also located to the vasculature thus overlapping with the endostatin staining. There was an increase in the expression rate of MMP-1 and MMP-2 in pancreas cancer, and although less data support the involvement of these MMPs in endostatin processing they most likely do influence the breakdown of the tissue ECM and BMs thus affecting the levels of other pro- and antiangiogenic substances.21,32 The negative effect of MMP inhibition on the release of antiangiogenic substances from the ECM could potentially be one explanation of the rather low success rate of this therapy as a pancreas cancer therapeutic in several clinical trials.33 We thus show that the expression pattern of type XVIII collagen/endostatin is shifted from a general BM pattern towards a VBM one when comparing normal and malignantly transformed pancreas. This shift also coincides with an upregulation of MMPs that have previously been shown to be involved in endostatin processing. Increased circulating levels of endostatin in patients with pancreas cancer and the effect of surgery and intraperitoneal chemotherapy As endostatin appears to be efficiently cleaved from type XVIII collagen in pancreas cancer we subsequently wanted to analyze the circulating levels of endostatin in patients with pancreas cancer. Circulating endostatin levels were analyzed from blood samples collected from patients with pancreas cancer as well as patients admitted for a nonmalignant disease (Table I). Most interestingly, we could observe that patients with pancreas cancer have a statistically significantly increased level of circulating endostatin when compared to controls (123 6 30 ng/ml vs. 76 6 17 ng/ml, p 5 0.006) (Fig. 3). It has previously been shown that patients with several cancers such as colorectal, breast, ovarian, renal, headand neck cancer, non-Hodgkin lymphoma and soft tissue sarcomas have increased levels of circulating endostatin.34–40 In many of these cancers an increased circulating endostatin level correlates with a poor prognosis. Although endostatin is a potent antiangiogenic substance the increase in patients with advanced cancer

FIGURE 3 – Circulating endostatin levels in patients with pancreas cancer. (a) Patients with pancreas cancer show significantly increased levels (p 5 0.006) of circulating endostatin when compared to the control patients. After treatment with either surgery or intraperitoneal chemotherapy the endostatin levels are significantly reduced (p 5 0.02 and p 5 0.03, respectively) and are not statistically significantly different than that of the control patients (p 5 0.14 and 0.5, respectively). (b) Pre and postoperative circulating endostatin levels in patients. The circulating endostatin level is reduced in all patients.

most likely is a reflection of tumor load and a general activation of the angiogenic milieu, as patients simultaneously often also display increased circulating levels of proangiogenic substances such as VEGF.37,41,42 Therefore, although having increased circulating levels of antiangiogenic molecules most likely the angiogenic balance is nevertheless tilted towards proangiogenesis and thus generation of a tumor neovasculature.43 The same finding has also been demonstrated in mice. Mice bearing a tumor have a higher circulating endostatin level than control mice.9 However this increase in endostatin is not enough to cause the angiogenic bal-

2809

ENDOSTATIN IN PANCREAS CANCER

ance to be switched towards antiangiogenesis. However, transgenic mice over-expressing endostatin and bearing the same tumor display reduced tumor growth, as the circulating endostatin level is high enough to push the angiogenic balance towards antiangiogenesis.9 We next wanted to study how treatment of pancreas cancer with surgery or intraperitoneal 5-FU chemotherapy affects the circulating endostatin levels. As normal tissue regeneration has been shown to effect circulating endostatin levels we chose to analyze the circulating levels at least 8 weeks after surgery and 12 weeks after the treatment with IP chemotherapy was initiated. Both forms of treatment led to a normalization and statistically significant reduction (68 6 40 ng/ml, p 5 0.02 in patients receiving surgical treatment and 89 6 14 ng/ml, p 5 0.03 in patients receiving chemotherapy) of circulating endostatin levels when compared to levels before treatment (Fig. 3). Most likely this reflects the consequence of tumor load reduction in the patients. However, the results indicate that measuring the circulating endostatin level could potentially be used as a biomarker for pancreas cancer progression. The results should be confirmed in larger patient materials. It will also be of importance to analyze whether an increase in circulating endostatin levels can be observed before disease relapse.

Concluding remarks The results from this study show that the expression pattern of type XVIII collagen/endostatin is shifted from a general BM staining pattern towards a VBM staining pattern in pancreas cancer when compared to normal pancreas tissue. This shift in expression pattern coincides with an upregulation of MMPs involved in endostatin processing and with increased circulating levels of endostatin in patients with pancreas cancer. As patients undergo treatment with either curative (surgery) or palliative (intraperitoneal 5-FU chemotherapy) intention the circulating endostatin levels are normalized. The potential use of endostatin as a biomarker for pancreas cancer progression needs to be studied further in a larger patient material. Acknowledgements This work was supported by the Finnish Medical Society Duo€ decim (M.S.), Cancerforskningsfonden Norrland (M.S., M.O., P.N.), Svenska S€allskapet f€ or Medicinsk Forskning SSMF (M.S.), V€asterbotten County Council and Insamlingsstiftelsen f€ or medicinsk forskning at Ume a˚ University (M.S.). The excellent technical assistance of Anette Berglund is gratefully acknowledged.

References 1.

2.

3.

4. 5. 6.

7.

8.

9.

10. 11. 12. 13. 14.

15.

Saarela J, Rehn M, Oikarinen A, Autio-Harmainen H, Pihlajaniemi T. The short and long forms of type XVIII collagen show clear tissue specificities in their expression and location in basement membrane zones in humans. Am J Pathol 1998;153:611–26. O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, Flynn E, Birkhead JR, Olsen BR, Folkman J. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell 1997;88:277– 85. Heljasvaara R, Nyberg P, Luostarinen J, Parikka M, Heikkila P, Rehn M, Sorsa T, Salo T, Pihlajaniemi T. Generation of biologically active endostatin fragments from human collagen XVIII by distinct matrix metalloproteases. Exp Cell Res 2005;307:292–304. Wen W, Moses MA, Wiederschain D, Arbiser JL, Folkman J. The generation of endostatin is mediated by elastase. Cancer Res 1999;59:6052–6. Felbor U, Dreier L, Bryant RA, Ploegh HL, Olsen BR, Mothes W. Secreted cathepsin L generates endostatin from collagen XVIII. Embo J 2000;19:1187–94. Lin HC, Chang JH, Jain S, Gabison EE, Kure T, Kato T, Fukai N, Azar DT. Matrilysin cleavage of corneal collagen type XVIII NC1 domain and generation of a 28-kDa fragment. Invest Ophthalmol Vis Sci 2001;42:2517–24. Kim YM, Jang JW, Lee OH, Yeon J, Choi EY, Kim KW, Lee ST, Kwon YG. Endostatin inhibits endothelial and tumor cellular invasion by blocking the activation and catalytic activity of matrix metalloproteinase. Cancer Res 2000;60:5410–3. Ma L, Elliott SN, Cirino G, Buret A, Ignarro LJ, Wallace JL. Platelets modulate gastric ulcer healing: role of endostatin and vascular endothelial growth factor release. Proc Natl Acad Sci USA 2001;98:6470– 5. Sund M, Hamano Y, Sugimoto H, Sudhakar A, Soubasakos M, Yerramalla U, Benjamin LE, Lawler J, Kieran M, Shah A, Kalluri R. Function of endogenous inhibitors of angiogenesis as endotheliumspecific tumor suppressors. Proc Natl Acad Sci USA 2005;102: 2934–9. Kalluri R. Basement membranes: structure, assembly and role in tumour angiogenesis. Nat Rev 2003;3:422–33. Folkman J. Endogenous angiogenesis inhibitors. APMIS 2004;112: 496–507. Dhanabal M, Volk R, Ramchandran R, Simons M, Sukhatme VP. Cloning, expression, and in vitro activity of human endostatin. Biochem Biophys Res Commun 1999;258:345–52. Dhanabal M, Ramchandran R, Waterman MJ, Lu H, Knebelmann B, Segal M, Sukhatme VP. Endostatin induces endothelial cell apoptosis. J Biol Chem 1999;274:11721–6. Karumanchi SA, Jha V, Ramchandran R, Karihaloo A, Tsiokas L, Chan B, Dhanabal M, Hanai JI, Venkataraman G, Shriver Z, Keiser N, Kalluri R, et al. Cell surface glypicans are low-affinity endostatin receptors. Mol Cell 2001;7:811–22. Rehn M, Veikkola T, Kukk-Valdre E, Nakamura H, Ilmonen M, Lombardo C, Pihlajaniemi T, Alitalo K, Vuori K. Interaction of endostatin

16.

17.

18. 19. 20. 21. 22.

23.

24.

25. 26. 27. 28. 29.

30.

with integrins implicated in angiogenesis. Proc Natl Acad Sci USA 2001;98:1024–9. Sudhakar A, Sugimoto H, Yang C, Lively J, Zeisberg M, Kalluri R. Human tumstatin and human endostatin exhibit distinct antiangiogenic activities mediated by a v b 3 and a 5 b 1 integrins. Proc Natl Acad Sci USA 2003;100:4766–71. Wickstrom SA, Alitalo K, Keski-Oja J. Endostatin associates with integrin a5b1 and caveolin-1, and activates Src via a tyrosyl phosphatase-dependent pathway in human endothelial cells. Cancer Res 2002;62:5580–9. Wickstrom SA, Alitalo K, Keski-Oja J. An endostatin-derived peptide interacts with integrins and regulates actin cytoskeleton and migration of endothelial cells. J Biol Chem 2004;279:20178–85. Rundhaug JE. Matrix metalloproteinases and angiogenesis. J Cell Mol Med 2005;9:267–85. Moses MA. The regulation of neovascularization of matrix metalloproteinases and their inhibitors. Stem Cells 1997;15:180–9. Werb Z, Vu TH, Rinkenberger JL, Coussens LM. Matrix-degrading proteases and angiogenesis during development and tumor formation. APMIS 1999;107:11–18. Hamano Y, Zeisberg M, Sugimoto H, Lively JC, Maeshima Y, Yang C, Hynes RO, Werb Z, Sudhakar A, Kalluri R. Physiological levels of tumstatin, a fragment of collagen IV a3 chain, are generated by MMP-9 proteolysis and suppress angiogenesis via aV b3 integrin. Cancer Cell 2003;3:589–601. Kamphaus GD, Colorado PC, Panka DJ, Hopfer H, Ramchandran R, Torre A, Maeshima Y, Mier JW, Sukhatme VP, Kalluri R. Canstatin, a novel matrix-derived inhibitor of angiogenesis and tumor growth. J Biol Chem 2000;275:1209–15. Colorado PC, Torre A, Kamphaus G, Maeshima Y, Hopfer H, Takahashi K, Volk R, Zamborsky ED, Herman S, Sarkar PK, Ericksen MB, Dhanabal M, et al. Anti-angiogenic cues from vascular basement membrane collagen. Cancer Res 2000;60:2520–6. Brammer RD, Bramhall SR, Eggo MC. Endostatin expression in pancreatic tissue is modulated by elastase. Br J Cancer 2005;92:89– 93. Brammer RD, Bramhall SR, Eggo MC. Endostatin expression in a pancreatic cell line is modulated by a TNFa-dependent elastase. Br J Cancer 2005;93:1024–8. Abdollahi A, Hahnfeldt P, Maercker C, Grone HJ, Debus J, Folkman J, Hlatky L, Huber PE. EndostatinÕs anti-angiogenic signaling network. Mol Cell 2004;13:649–63. Oman M, Blind PJ, Naredi P, Gustavsson B, Hafstrom LO. Treatment of non-resectable pancreatic cancer with intraperitoneal 5-FU and leucovorin IV. Eur J Surg Oncol 2001;27:477–81. Oman M, Lundqvist S, Gustavsson B, Hafstrom LO, Naredi P. Phase I/II trial of intraperitoneal 5-Fluorouracil with and without intravenous vasopressin in non-resectable pancreas cancer. Cancer Chemother Pharmacol 2005;56:603–9. Jones LE, Humphreys MJ, Campbell F, Neoptolemos JP, Boyd MT. Comprehensive analysis of matrix metalloproteinase and tissue inhibi-

2810

31. 32. 33. 34.

35.

36. 37.

€ OHLUND ET AL.

tor expression in pancreatic cancer: increased expression of matrix metalloproteinase-7 predicts poor survival. Clin Cancer Res 2004; 10:2832–45. Bodey B, Bodey B, Jr, Siegel SE, Kaiser HE. Immunocytochemical detection of the expression of members of the matrix metalloproteinase family in adenocarcinomas of the pancreas. In Vivo 2001;15:71–6. Ferreras M, Felbor U, Lenhard T, Olsen BR, Delaisse J. Generation and degradation of human endostatin proteins by various proteinases. FEBS Lett 2000;486:247–51. Saif MW. Anti-angiogenesis therapy in pancreatic carcinoma. JOP 2006;7:163–73. Feldman AL, Alexander HR, Jr, Bartlett DL, Kranda KC, Miller MS, Costouros NG, Choyke PL, Libutti SK. A prospective analysis of plasma endostatin levels in colorectal cancer patients with liver metastases. Ann Surg Oncol 2001;8:741–5. Feldman AL, Alexander HR, Jr, Yang JC, Linehan WM, Eyler RA, Miller MS, Steinberg SM, Libutti SK. Prospective analysis of circulating endostatin levels in patients with renal cell carcinoma. Cancer 2002;95:1637–43. Feldman AL, Pak H, Yang JC, Alexander HR, Jr, Libutti SK. Serum endostatin levels are elevated in patients with soft tissue sarcoma. Cancer 2001;91:1525–9. Zhao J, Yan F, Ju H, Tang J, Qin J. Correlation between serum vascular endothelial growth factor and endostatin levels in patients with breast cancer. Cancer Lett 2004;204:87–95.

38. Bono P, Teerenhovi L, Joensuu H. Elevated serum endostatin is associated with poor outcome in patients with non-Hodgkin lymphoma. Cancer 2003;97:2767–75. 39. Hata K, Dhar DK, Kanasaki H, Nakayama K, Fujiwaki R, Katabuchi H, Okamura H, Nagasue N, Miyazaki K. Serum endostatin levels in patients with epithelial ovarian cancer. Anticancer Res 2003;23: 1907–12. 40. Homer JJ, Greenman J, Stafford ND. Circulating angiogenic cytokines as tumour markers and prognostic factors in head and neck squamous cell carcinoma. Clin Otolaryngol 2002;27:32–7. 41. Galizia G, Ferraraccio F, Lieto E, Orditura M, Castellano P, Imperatore V, Romano C, Vollaro M, Agostini B, Pignatelli C, De Vita F. Prognostic value of p27, p53, and vascular endothelial growth factor in Dukes A and B colon cancer patients undergoing potentially curative surgery. Dis Colon Rectum 2004; 47:1904–14. 42. De Vita F, Orditura M, Lieto E, Infusino S, Morgillo F, Martinelli E, Castellano P, Romano C, Ciardiello F, Catalano G, Pignatelli C, Galizia G. Elevated perioperative serum vascular endothelial growth factor levels in patients with colon carcinoma. Cancer 2004; 100:270–8. 43. Sund M, Zeisberg M, Kalluri R. Endogenous stimulators and inhibitors of angiogenesis in gastrointestinal cancers: basic science to clinical application. Gastroenterology 2005;129:2076–91.