Angiogenesis, Lymphangiogenesis, and Melanoma Metastasis

Angiogenesis, lymphangiogenesis, and melanoma metastasis

Michael Streit1 and Michael Detmar*,2

1Berlex Biosciences, Richmond, CA 94804, USA; 2Cutaneous Biology Research Center and Department of Dermatology, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA 02129, USA

The induction of angiogenesis is a critical point in the interplay between environmental and genetic factors development of most human tumors – including melano- determines the balance and activity of pro- and mas. Some of the earliest studies in the field of tumor antiangiogenic factors, which may be specific for distinct angiogenesis showed that transplantation of human tumor types and tumor locations, and which may also melanoma fragments into the hamster cheek pouch be altered during tumor progression (Folkman, 2002). stimulated blood vessel growth. Since then, numerous Recent findings have also indicated that lymphangio- studies have demonstrated that human melanomas also genesis might be involved in melanoma metastasis induce angiogenesis. The prognostic importance of the (Dadras, 2003). But what is the clinical and experi- degree of melanoma vascularization, however, has re- mental evidence that angiogenesis is an important aspect mained controversial. Elevated expression of several of melanoma progression, and will antiangiogenic angiogenic factors, including vascular endothelial growth therapies be useful in treating these patients? factor, basic fibroblast growth factor, and interleukin-8, has been detected in primary cutaneous melanomas, and the importance of these mediators in promoting melanoma Tumor angiogenesis in melanoma angiogenesis and metastasis has been confirmed in tumor xenotransplant models. Based upon these findings, several New blood vessel formation is a prominent feature of clinical trials of angiogenesis inhibitors have been initiated human cutaneous melanomas, indicating that these in human melanoma patients and are currently underway. tumors have angiogenic activity (Mihm et al., 1975). Recent experimental evidence indicates that tumor- The observation that cutaneous melanoma cells acquire associated lymphangiogenesis also plays an important the capacity to actively induce the growth of new blood role in mediating tumor spread to regional lymph nodes. vessels dates back to the earliest days of tumor angiogen- These observations have important implications for esis research (Warren and Shubik, 1966; Hubler and Wolf, prognosis and treatment of human melanomas. 1976; Stenzinger et al., 1983). Warren and Shubick (1966) Oncogene (2003) 22, 3172–3179. doi:10.1038/sj.onc.1206457 first observed the induction of tumor angiogenesis after transplantation of human melanoma fragments into the Keywords: VEGF; FGF; LYVE-1; VEGF-C hamster cheek pouch; these results were confirmed in later studies (Hubler and Wolf, 1976). Using Doppler ultra- sound studies, Srivastava et al. (1986) detected tumor Introduction blood flow in most melanomas that were more than The induction of angiogenesis – the generation of new 0.9 mm thick, but rarely in thinner tumors. The clinical capillary blood vessels from pre-existing vessels – is and prognostic significance of tumor angiogenesis for generally considered as essential to ensure the supply of melanoma progression and metastasis, however, has oxygen and nutrients for malignant tumor growth, remained controversial. In an initial study of 20 melano- et al invasion, and metastasis (Hanahan and Weinberg, mas, Srivastava . (1988) reported that the vascular 2000). Whereas blood vessel growth is tightly controlled area at the base of melanomas with locoregional or under physiological conditions, tumor progression is systemic metastases was more than twice that of matched frequently associated with the acquisition of an angio- recurrence-free tumors. Subsequent studies, using the Ulex europaeus genic phenotype, associated with a switch in the balance lectin agglutinin I to detect blood vessels, of pro- and antiangiogenic molecules (Hanahan and showed a gradual rise in vascularity as melanocytic Folkman, 1996). Multiple signals trigger the angiogenic tumors progressed. Onset of angiogenesis occurred during switch, including genetic mutations, metabolic and the radial growth phase of cutaneous melanomas, mechanical stress and inflammatory response mechan- although considerable heterogeneity was noted between et al isms (Carmeliet and Jain, 2000). Moreover, a complex individual cases (Barnhill ., 1992). Moreover, the onset of angiogenesis in thin melanomas was associated with inflammatory regression and with development of *Correspondence: M Detmar, Cutaneous Biology Research Center and Department of Dermatology, Massachusetts General Hospital, the vertical growth phase (Barnhill and Levy, 1993). Building 149, 13th Street, Charlestown, MA 02129, USA; A rise in the mean vascular density was correlated E-mail: [email protected] with melanoma progression in subsequent histochemical Angiogenesis and melanoma M Streit and M Detmar 3173 studies of cutaneous melanomas (Erhard et al., 1997; Table 1 Differential expression of vascular markers in cutaneous Marcoval et al., 1997), and a number of retrospective lymphatic and blood vessels histological studies have reported an inverse correlation Marker Lymphatics Blood vessels between tumor microvessel density and disease-free and CD34 À + overall survival of melanoma patients (Srivastava et al., CD44 À + 1988; Graham et al., 1994; Straume et al., 1999; PAL-E À + Vlaykova et al., 1999). A recently published prospective Collagen type IV À/(+) + study of 417 cutaneous melanoma patients reported that Collagen type XVIII À/(+) + tumor vascularity, as determined by analysis of routine Laminin À/(+) ++ Neuropilin-1 À + histologic stains, was the most important determinant of overall survival, surpassing tumor thickness (Kashani- VEGFR-3 + À Sabet et al., 2002). Several other investigators, in Podoplanin + À contrast, failed to detect any correlation between SLC/CCL21 + À LYVE-1 + À melanoma vascularization and prognosis. Prox1 + À Using morphometric analysis of Ulex europaeus type I lectin-labeled sections obtained from 86 melanomas CD31 (PECAM-1) + ++ with no evidence of recurrence after a minimum follow- VEGFR-2 + + up period of 5 years and of 21 cases with locoregional Factor VIII-related antigen + ++ recurrence and/or metastasis, Carnochan et al. (1991) found that tumor recurrence could not be predicted by any of the derived vascular parameters (vascular length, benign premalignant and malignant epithelial experi- surface, and volume density) – either independently or mental and human skin tumors (Weninger et al., 1996; when combined with other histological and clinical Hawighorst et al., 2001, 2002). features. Busam et al. (1995) compared 60 cases of Importantly, none of the studies published on metastasizing and nonmetastasizing cutaneous melano- melanoma blood vessel quantification have included mas that were matched for tumor thickness, age, sex, recently discovered molecular markers that can be used and anatomic site. In this study, there was no significant to specifically detect blood vessels and lymphatic vessels difference in the number of microvessels or in the in tissue sections (Detmar and Hirakawa, 2002; Oliver pattern of vascular microarchitecture between metasta- and Detmar, 2002) (Table 1). Additional studies should sizing and nonmetastasizing tumors. Recent computer- be performed on primary tumor samples taken from assisted image analysis of matched cases of 20 non- matched cohorts of melanoma patients, with long-term metastatic and 20 metastatic cutaneous melanomas follow-up, to determine whether angiogenesis is a useful revealed a comparable level of vascularization in both prognostic factor for this cancer type. These studies tumor types, as determined by immunohistochemical should also include the most recently developed, highly quantification of vessels stained for the vascular junction specific and sensitive methods for detecting tumor- molecule CD31 (PECAM-1) (M Detmar, unpublished associated blood vessels. results). More surprisingly, a similar immunohistochemical analysis of primary cutaneous melanomas samples taken from 84 patients showed that the only indepen- Expression of angiogenesis modulators by human dent variable associated with disease-free survival was melanomas tumor thickness (Breslow classification). The only variable associated with overall survival was depth of Multiple proangiogenic factors are produced by primary tumor infiltration (Clark classification), whereas high cutaneous melanoma cells. Vascular endothelial growth levels of vascularity were actually correlated with factor (VEGF), also known as vascular permeability survival (Ilmonen et al., 1999). factor (Senger et al., 1983), is an endothelial cell-specific Thus, the potential prognostic value of tumor growth factor and the principal regulator of angiogen- vascularization in human cutaneous melanomas remains esis under normal and pathological conditions in most unresolved. Some of these contradictory results might be organs (Dvorak et al., 1995; Ferrara, 1999) including the explained by the nonstandardized assessments of tumor skin (Detmar, 1996). Using immunohistochemical ana- vascularity – these include ‘hot spot’ analysis and lyses, the transition of horizontal to vertical growth representative field analysis – and by the various phase in melanoma was found to be associated with detection methods used to visualize tumor-associated increased VEGF protein expression and accumulation blood vessels (routine histologic stains versus vascular in the tumor stroma (Erhard et al., 1997; Marcoval et al., immunostains). As an example, Factor VIII-related 1997). Immunoreactivity for VEGF was also found to antigen has been frequently used to detect tumor vessels, be related to melanoma thickness, although VEGF was but this protein is only expressed by a fraction of not found to be a useful prognostic indicator for immature, CD31-positive intratumoral blood vessels; so malignant melanoma (Bayer-Garner et al., 1999; quantification techniques based on the presence of this Straume and Akslen, 2001). Other studies showed that marker might not include all tumor-associated blood VEGF was only expressed in 32% of primary melano- vessels. Moreover, recent evidence has shown that the mas, with increased expression levels in metastases extent of vascularization does not discriminate between (Salven et al., 1997; Vlaykova et al., 1999). Although

Oncogene Angiogenesis and melanoma M Streit and M Detmar 3174 these studies demonstrated that VEGF expression is not these mutations, coupled with decreased TSP-1 expres- as prominent in melanomas as in most epithelial cancers sion, might promote the metastatic phenotype in (Dvorak et al., 1995) and, therefore, might not represent malignant melanoma (Grant et al., 1998). the major angiogenic activity in these tumors, the expression of functional VEGF receptors on human melanoma cells suggests the intriguing possibility that Angiogenesis in experimental melanoma models VEGF might also exert autocrine effects on the tumor cells themselves (Gitay-Goren et al., 1993; Liu et al., Human melanoma cells synthesize a plethora of 1995; Graeven et al., 1999; Lacal et al., 2000). angiogenic factors in vitro, including VEGF, bFGF, Expression of mRNA for basic fibroblast growth IL-8, PDGF, and PlGF. The specific biological function factor (bFGF, FGF-2), a potent angiogenesis factor, has of several of these factors has been evaluated in both in been detected in metastatic and primary invasive vitro angiogenesis models and in xenograft models. melanomas, whereas melanocytes in benign nevi did VEGF expression was found to be absent in normal not express this factor (Reed et al., 1994) . Melanomas melanocytes but upregulated in malignant melanoma that were undergoing regression and dysplastic nevi cells cells (Gitay-Goren et al., 1993). In a series of melanoma were also observed to express bFGF mRNA, however, cell lines, in vitro VEGF expression was correlated with and bFGF expression levels were also found to be the degree of tumor angiogenesis and the metastatic higher in benign melanocytic nevi than in malignant potential of in vivo tumor xenografts (Potgens et al., melanoma cells (Ahmed et al., 1997). It has even been 1995). Overexpression of VEGF in SK-MEL-2 melano- reported that melanomas with bFGF-positive vessels ma cells, which normally express only low baseline had a better prognosis when compared with melanomas VEGF levels, promoted tumor growth, angiogenesis, with bFGF-negative vessels (Straume and Akslen, and metastasis in vivo (Claffey et al., 1996). 2002), suggesting that bFGF upregulation cannot be VEGF is a homodimeric, heparin-binding glycopro- correlated with malignant transformation in human tein occurring in at least four isoforms of 121, 165, 189, melanomas. The recent observation, however, that low and 201 amino acids because of alternative splicing. The bFGF serum concentrations at the beginning of high- different effects of the various VEGF isoforms were dose interferon (IFN)-a2b therapy were associated with recently studied in the nontumorigenic human WM1341 recurrence-free survival (Dreau et al., 2001), together B melanoma cell line, which was derived from early stage with the results of experimental tumor xenotransplant melanoma cells that do not express VEGF (Yu et al., studies (see below), warrants further investigation into 2002). Overexpression of VEGF121 and of VEGF165 by the importance of bFGF in melanoma progression. these cells resulted in aggressive tumor growth as mouse Several other angiogenic factors have been implicated xenografts, whereas cells that overexpressed VEGF189 in the pathology of human melanomas. Interleukin (IL)- remained nontumorigenic and dormant after injection 8, in particular, was found to be absent from normal into mice. This difference was likely to be because of the epidermis and benign melanocytic lesions but was reduced bioavailability of the VEGF189 isoform, which expressed at high levels in the majority of cutaneous has high heparin affinity and remains cell associated after melanomas examined (Nurnberg et al., 1999). Moreover, secretion. The VEGF165-expressing tumors developed a in a series of 125 melanoma patients, IL-8 serum levels much denser blood vessel network than tumors that were found to be elevated, compared to healthy controls, overexpressed VEGF121, which is the predominant and were correlated with an advanced disease stage and isoform detected in human melanomas. poor progression-free and overall survival (Ugurel et al., These findings have important implications for the 2001). Immunohistochemical analysis revealed increased pattern of tumor vascularization that occurs during expression levels of the angiogenic factors placental melanoma progression (Yu et al., 2002). In addition to growth factor (PlGF; Lacal et al., 2000), platelet-derived the tumor-type specific expression pattern and activity of growth factor (PDGF)-AA and – BB (Barnhill et al., different VEGF isoforms, recent results provide evidence 1996), and angiogenin (Hartmann et al., 1999) in human that the angiogenic response to a distinct melanoma- melanoma tissue samples. However, the prognostic value derived VEGF isoform might be site specific. When SK- of these factors and their potential importance for MEL melanoma cells that overexpressed murine melanoma metastasis remain at present unclear. VEGF164 (the equivalent isoform of human VEGF165) Downregulation of endogenous angiogenesis inhibi- were subcutaneously implanted in mice, they induced tors has been observed in several epithelial cancers, and blood vessel sprouting of new blood vessels (Claffey has been proposed to enhance tumor progression. In et al., 1996). On the other hand, brain metastases derived contrast to the large number of studies of proangiogenic from the human melanoma cell line Mel57 that over- molecules, however, little is known about the expression expressed VEGF165 co-opted pre-existing intra- and of endogenous angiogenesis inhibitors by melanomas. peritumoral vessels without inducing angiogenic sprout- An inverse correlation between mutations in the tumor ing of new vessels (Kusters et al., 2002). suppressor p53 and the expression of the angiogenesis Basic FGF is expressed by most melanoma cells, but inhibitor thrombospondin (TSP)-1 was found in a study not by normal melanocytes, whereas both types of cells of a series of 99 melanoma samples. Researchers express high-affinity receptors for bFGF (Lazar-Molnar detected a significantly higher incidence of p53 muta- et al., 2000). Although overexpression of bFGF tions in metastatic tumors, suggesting that acquisition of conferred the capacity for anchorage-independent

Oncogene Angiogenesis and melanoma M Streit and M Detmar 3175 growth to human and murine melanocytes in vitro, Recently, a highly patterned system of vascular bFGF-transfected melanocytes did not form persisting channels that were devoid of endothelial cells but were malignant tumors in vivo, indicating that autocrine instead composed of extracellular matrix, lined exter- bFGF stimulation provides a growth advantage but is nally by tumor cells, was observed in some aggressive not sufficient for the induction of a transformed human uveal and cutaneous melanomas (Maniotis et al., phenotype (Dotto et al., 1989; Nesbit et al., 1999; 1999). This feature, termed ‘vasculogenic mimicry’, was Graeven et al., 2001). Interference with the bFGF also observed after injection of melanoma cells into an pathway in melanoma cells by antisense approaches and ischemic microenvironment that was surgically induced overexpression of a dominant-negative mutant FGF in the hind limbs of nude mice (Hendrix et al., 2002). receptor 1 was shown to inhibit in vitro cell proliferation This formation of tubular networks could also be and in vivo tumorigenesis (Becker et al., 1989, 1992; created in vitro by growing melanoma cells in three- Yayon et al., 1997), whereas overexpression of bFGF dimensional cultures (Seftor et al., 2001). The implica- promoted the in vivo tumor growth and tumor tions of these findings for the pathogenesis and angiogenesis of WM164 melanoma cells (Graeven progression of human melanomas are at present unclear. et al., 2001). In these studies, downregulation of VEGF expression slowed tumor growth, whereas transfection of a bFGF antisense construct completely inhibited Therapeutic targeting of angiogenesis in melanoma tumor formation, indicating an important autocrine patients function of bFGF in human melanoma development. IL-8 and PDGF have also been implicated in the More than two decades ago, Folkman (1975) proposed promotion of experimental melanoma growth, angio- that inhibition of tumor angiogenesis might represent a genesis, and metastasis (Rofstad and Halsor, 2000). IL-8 new strategy for treating human cancers. Increasing is produced by melanocytes and melanoma cells. It experimental evidence, obtained predominantly in tu- stimulates cell migration, proliferation, and metastasis mor xenotransplant models, suggests that in addition to in an autocrine fashion, as shown in IL-8-transfected epithelial cancers, malignant melanoma growth and nonmetastatic SB-2 melanoma cells (Luca et al., 1997). progression might also be inhibited by blockade of Moreover, in D-12 human melanoma xenografts, the blood vessel growth (Table 2). Several reagents have incidence of spontaneous metastasis was associated with been developed to block VEGF activity, including increased IL-8 expression, and treatment with IL-8 neutralizing antibodies against VEGF, fusion proteins neutralizing antibodies significantly decreased angiogen- of the VEGF121 isoform and diphteria toxin, and small esis and metastasis formation (Rofstad and Halsor, molecules or antibodies that prevent VEGF from 2002). In human WM9 melanoma cells that do not binding to or signaling through its receptors on the express PDGF, induction of tumor-associated blood vascular endothelium. These have been demonstrated to vessels and formation of a dense connective tissue have antitumor and antiangiogenic activity in melanoma stroma were observed after cells were transfected with a models (Yuan et al., 1996; Ferrara and Alitalo, 1999; PDGF expression vector (Forsberg et al., 1993). Wild et al., 2000). Moreover, studies designed to block

Table 2 Expression patterns and prognostic relevance of angiogenic factors in malignant melanoma Angiogenic factor Technique Findings References VEGF RT–PCR, No expression in melanocytes, benign or dysplastic nevi Graeven et al. (1999) IHC IHC Strong expression in melanoma and accumulation in tumor Erhard et al. (1997) stroma IHC Melanoma expression correlates with tumor thickness and Marcoval et al. (1997) absence of regression IHC Melanoma expression correlates with tumor thickness and Bayer-Garner et al. (1999) absence of regression IHC Increased expression in less vascularized melanoma lesions Straume and Akslen (2001) IHC Melanoma expression correlates with the occurrence of organ Salven et al. (1997) metastases IHC Melanoma expression correlates with increased vascular density Vlaykova et al. (1999) bFGF ISH No expression in benign nevi and melanoma in situ; expression Read et al. (1994) in dysplastic nevi and invasive melanoma IHC Heterogeneous expression in malignant melanoma Ahmed et al. (1997) IHC Melanoma expression correlates with microvascular density; Straume and Akslen (2002) expression in tumor vasculature correlates with improved prognosis

IL-8 ISH No expression in melanocytic nevi; melanoma expression Nurnberg et al. (1999) correlates with superﬁcial-spreading subtype and reduced time to disease progression

RT–PCR, Reverse transcriptase polymerase chain reaction; IHC, immunohistochemistry; ISH, in situ hybridization

Oncogene Angiogenesis and melanoma M Streit and M Detmar 3176 bFGF activity, such as by vaccination of animals with (Posey et al., 2001), and the angiogenesis inhibitor the heparin-binding domain of bFGF, have been shown thalidomide, combined with the cytotoxic agent temo- to inhibit tumor angiogenesis, melanoma growth, and/ zolomide (Hwu et al., 2002), are currently being tested in or metastasis (Becker et al., 1989; 1992; Plum et al., clinical trials for patients with metastatic melanoma. 2000; Yayon et al., 1997). Other antiangiogenic strate- The combination of temozolomide and thalidomide was gies that have shown antitumoral efficacy in preclinical shown to be well tolerated and had antitumor activity in melanoma models include blockade/targeting of the some patients with advanced melanoma (Hwu et al., urokinase-type plasminogen activator (Min et al., 1996), 2002). A phase I trial of the angiogenesis inhibitor TNP- Tie-2 receptor ligands (Lin et al., 1998), the av integrin 470, a derivative of fumagillin, resulted in the induction receptor (Lode et al., 1999), and matrix metalloprotei- of long-term stable disease in one patient with progres- nases (Naglich et al., 2001). Moreover, overexpression sive metastatic melanoma (Bhargava et al., 1999). Small or systemic application of the endogenous angiogenesis molecule inhibitors that target VEGF, FGF, or PDGF inhibitors angiostatin (O’Reilly et al., 1994), endostatin tyrosine kinase receptors have also shown some (Boehm et al., 1997; O’Reilly et al., 1997), TSP-1 (Miao promising results in phase I/II studies of several et al., 2001; Reiher et al., 2002), TSP-2 (Streit et al., different types of cancer patients (Morin, 2000). The 2002), and IL-10 (Huang et al., 1996) has been shown to efficacy of these reagents in human patients with slow tumor growth in melanoma xenograft models. melanoma, however, is unclear. Since malignant mela- Recent findings suggest that tumor vasculature noma cells release a number of different proangiogenic expresses a specific set of endothelial cell genes which factors, a combination of these and other antiangiogenic differ from those expressed by normal nonproliferating agents, combined with traditional or low-dose che- blood vessels (St Croix et al., 2000). Although the motherapy or with immunotherapy, might ultimately be expression of many of these genes is shared by needed to inhibit melanoma growth. angiogenic vessels in nonmalignant conditions, the identification of markers that are specific for proliferat- ing microvascular tumor endothelial cells might lead to Tumor lymphangiogenesis and metastasis new therapeutics designed to specifically target tumor- associated blood vessels. However, it is important to Tumor metastasis to regional lymph nodes, as detected remember that the long-term efficacy of antiangiogenic by analysis of sentinel lymph nodes, frequently repre- agents might be limited by the complex biology of sents the first step of melanoma dissemination and aggressive malignant melanomas. Recently, subpopula- serves as an important prognostic indicator. In contrast tions of melanoma cells were shown to survive even to the extensive studies on melanoma-associated angio- under conditions of hypoxia and metabolic stress, genesis, however, little is known about the mechanisms indicating that they might be able to withstand angio- by which melanoma cells gain entry into the lymphatic genesis inhibition (Rak et al., 2002), and xenografts of system. The recent identification of lymphatic growth WM239A melanoma cells have been shown to have high factors (VEGF-C and VEGF-D) and their receptor levels of heterogeneity, with respect to vascular depen- (VEGFR-3), along with the discovery of lymphatic- dency and tolerance for hypoxia (Yu et al., 2001). specific markers (Table 1), has provided new tools for The treatment options for metastatic malignant studying the formation of tumor-associated lymphatic melanoma are limited and the prognosis for these vessels (Oliver and Detmar, 2002) and their contribution patients is poor. Currently, surgical removal of the to lymphatic tumor spread. primary tumor and of the draining regional lymph An increasing number of clinicopathological studies nodes remains the most effective treatment in patients have shown a direct correlation between tumor expres- with nodal disease, whereas systemic mono-, poly , or sion of the lymphangiogenesis factors VEGF-C or immunochemotherapy regimens have achieved only VEGF-D and metastatic tumor spread in many human marginal response rates without significant impact on cancers (reviewed in Stacker et al., 2002), providing patient survival (Cascinelli et al., 2000). Consequently, evidence for the involvement of lymphangiogenesis in early detection and prevention of disease progression tumor progression. Not much data are available, has remained the most efficient strategy to reduce however, regarding the expression of lymphangiogenesis melanoma mortality, and innovative therapies for factors by melanomas. A375 melanoma cells, which advanced human melanoma are urgently needed. Tumor express VEGF-C, were found to induce lymphatic vessel blood vessels represent an alternative therapeutic target growth and to invade lymphatic vessels in the avian for long-term treatment of melanoma, since tumor chorioallantoic membrane, whereas VEGF-C-negative progression appears to be angiogenesis dependent and Malme 3 M melanoma cells had no such effects (Papoutsi because vascular endothelial cells do not have the et al., 2000). In human melanomas, VEGF-C mRNA genetic instability that allows cancer cells to develop expression was detected by primary cutaneous melano- drug resistance (Carmeliet and Jain, 2000). mas (Salven et al., 1998). VEGF-D protein was detected The safety, feasibility and efficacy of antiangiogenic in tumor cells and in vessels that were adjacent to therapies for patients with advanced malignant melano- immunopositive tumor cells, but not in normal melanoma are currently under investigation in clinical phase I– cytes or in vessels that were distant from the tumors III trials (Table 3). The humanized monoclonal anti- (Achen et al., 2001). This indicates a paracrine regulation body Vitaxin, which is directed against the avb3 integrin of tumor lymphangiogenesis in some melanomas.

Oncogene Angiogenesis and melanoma M Streit and M Detmar 3177 Table 3 Anti-angiogenic therapy in malignant melanoma: clinical and experimental approaches Drug Molecule class Development stage References Thalidomide Small-molecule inhibitor Phase I–III clinical trials Eisen et al. (2000) Hwu et al. (2000) TNP470 Small-molecule inhibitor Phase I/II clinical trials Bhargava et al. (1999) Vitaxin Anti-avb3 integrin antibody Phase I/II clinical trials Posey et al. (2001) Target/drug Method Results Reference VEGF Neutralizing antibody Inhibition of xenograft growth Yuan et al. (1996) VEGF-121 Diphteria toxin fusion protein Inhibition of xenograft growth and metastasis Wild et al. (2000) VEGF-R2 Dendritic cell vaccination Inhibition of xenograft growth and metastasis Li et al. (2002) (ﬂk-1) IL-8 Neutralizing antibody Inhibition of xenograft growth Huang et al. (2002) uPA Neutralizing antibody Inhibition of xenograft growth Min et al. (1996) Tie-2 Soluble receptor fragments Inhibition of xenograft growth Lin et al. (1998) av-Integrin Short peptides Inhibition of xenograft growth Kumar et al. (2001) Buerkle et al. (2002 av-Integrin IL-2 fusion protein growth Inhibition of xenograft Lode et al. (1999) and metastasis MMP Small-molecule inhibitors Angiogenesis inhibition Naglich et al. (2001) FGF Peptide vaccination Inhibition of tumor angiogenesis, growth, Plum et al. (2000) and metastasis Angiostatin Recombinant protein Inhibition of xenograft growth O’Reilley (1994) Endostatin Recombinant protein Inhibition of xenograft growth O’Reilley (1997) Boehm et al. (1997) TSP-1 Recombinant proteins, Inhibition of angiogenesis and xenograft growth Miao et al. (2001) peptide mimetics Reiher et al. (2002) TSP-2 Experimental gene therapy Inhibition of angiogenesis and xenograft growth Streit (2002)

Recent studies in animal tumor models have provided cells. Lymphatic endothelium secretes the chemokine direct evidence that increased levels of VEGF-C or CCL21 (secondary lymphoid chemokine), which binds VEGF-D promote tumor lymphangiogenesis and lym- to CC chemokine receptor 7 (CCR7) (Gunn et al., 1998; phatic tumor spread to regional lymph nodes, and that Saeki et al., 1999), leading to chemoattraction and these effects can be suppressed by blocking VEGFR-3 migration of mature dendritic cells from the skin to signaling (Karpanen et al., 2001; Mandriota et al., 2001; regional lymph nodes. CCR7 is also expressed by some Skobe et al., 2001a, b; Stacker et al., 2001; He et al., malignant melanoma cell lines (Muller et al., 2001), and 2002). Whereas VEGF-C overexpression induced tumor overexpression of CCR7 in B16 malignant melanoma lymphangiogenesis and angiogenesis in MeWo melanoma cells led to a dramatic increase in the incidence of xenotransplants (Skobe et al., 2001a), the impact of regional lymph node metastases after injection into the enhanced levels of VEGF-C or VEGF-D on experimental footpad of mice (Wiley et al., 2001). These findings melanoma metastasis has not yet been studied. These indicate that some tumors might take advantage of studies have been made possible by the discovery of the molecular mechanisms designed for the physiological lymphatic endothelium-specific marker LYVE-1 – a immune response to further their metastatic spread. hyaluronan receptor with homology to CD44 (Prevo In conclusion, there is now much evidence to indicate et al., 2001). Antibodies against LYVE-1 can be used to that angiogenesis induction is an important part of identify and quantify tumor-associated lymphatic vessels. melanoma progression, although the significance of The specificity of LYVE-1 for tumor-associated lympha- melanoma vascularization as a prognostic indicator tic endothelium has been confirmed in a number of remains unclear. As an increasing number of angiogen- experimental tumors, in which all LYVE-1-positive esis inhibitors are being tested on other types of tumors, tumor-associated lymphatic vessels also expressed the these trials should be expanded to include melanoma lymphatic-specific transcription factor Prox1 (Hawighorst patients with progressive disease. Future studies are et al., 2002; Oliver and Detmar, 2002; Wigle et al., 2002). needed to investigate whether the expression of lym- Taken together, these results indicate that tumor- phangiogenesis factors and the extent of tumor-asso- associated lymphangiogenesis, which is induced by ciated lymphangiogenesis might serve as new prognostic VEGF-C, VEGF-D, or other unidentified growth tools for the evaluation of primary cutaneous melanoma. factors, can lead to proliferation and enlargement of peritumoral and intratumoral lymphatic vessels that have been detected in human melanomas (Oliver and Acknowledgements Detmar, 2002). The ability to promote lymphangiogen- This work was supported by NIH/NCI Grants CA69184, esis is likely to enhance the metastatic spread of CA86410 and CA91861 (MD), by American Cancer Society melanomas. Tumor-associated lymphatic endothelial ProgrAm. Project Grant 99-23901 (MD), and by the Cuta- cells are also likely to be involved in the chemotactic neous Biology Research Center through the Massachusetts recruitment and intralymphatic transport of melanoma General Hospital/Shiseido Co. Ltd Agreement (MD).

Oncogene Angiogenesis and melanoma M Streit and M Detmar 3178 References

Achen MG, Williams RA, Minekus MP, Thornton GE, Graeven U, Fiedler W, Karpinski S, Ergun S, Kilic N, Rodeck Stenvers K, Rogers PA, Lederman F, Roufail S and Stacker U, Schmiegel W and Hossfeld DK. (1999). J. Cancer Res. SA. (2001). J. Pathol., 193, 147–154. Clin. Oncol., 125, 621–629. Ahmed NU, Ueda M, Ito A, Ohashi A, Funasaka Y and Graeven U, Rodeck U, Karpinski S, Jost M, Philippou S and Ichihashi M. (1997). Melanoma Res., 7, 299–305. Schmiegel W. (2001). Cancer Res., 61, 7282–7290. Barnhill RL, Fandrey K, Levy MA, Mihm Jr MC and Hyman Graham CH, Rivers J, Kerbel RS, Stankiewicz KS and White B. (1992). Lab. Invest., 67, 331–337. WL. (1994). Am. J. Pathol., 145, 510–514. Barnhill RL and Levy MA. (1993). Am. J. Pathol., 143, Grant SW, Kyshtoobayeva AS, Kurosaki T, Jakowatz J and 99–104. Fruehauf JP. (1998). Cancer Detect Prev., 22, 185–194. Barnhill RL, Xiao M, Graves D and Antoniades HN. (1996). Gunn MD, Tangemann K, Tam C, Cyster JG, Rosen SD and Br. J. Dermatol., 135, 898–904. Williams LT. (1998). Proc. Natl. Acad. Sci. USA, 95, Bayer-Garner IB, Hough Jr AJ and Smoller BR. (1999). Mod. 258–263. Pathol., 12, 770–774. Hanahan D and Folkman J. (1996). Cell, 86, 353–364. Becker D, Lee PL, Rodeck U and Herlyn M. (1992). Hanahan D and Weinberg RA. (2000). Cell, 100, 57–70. Oncogene, 7, 2303–2313. Hartmann A, Kunz M, Kostlin S, Gillitzer R, Toksoy A, Becker D, Meier CB and Herlyn M. (1989). EMBO J., 8, Brocker EB and Klein CE. (1999). Cancer Res., 59, 1578– 3685–3691. 1583. Bhargava P, Marshall JL, Rizvi N, Dahut W, Yoe J, Figuera Hawighorst T, Oura H, Streit M, Janes L, Nguyen L, Brown M, Phipps K, Ong VS, Kato A and Hawkins MJ. (1999). LF, Jackson DG and Detmar M. (2002). 21, 7945–7956. Clin. Cancer Res., 5, 1989–1995. Hawighorst T, Velasco P, Streit M, Kyriakides TR, Brown Boehm T, Folkman J, Browder T and O’Reilly MS. (1997). LF, Bornstein P and Detmar M. (2001). EMBO J., 20, 2631– Nature, 390, 404–407. 2640. Buerkle MA, Pahernik SA, Sutter A, Jonczyk A, Messmer K, He Y, Kozaki K, Karpanen T, Koshikawa K, Yla-Herttuala S, Dellian M. (2002). Br. J. Cancer., 86, 788–795. Takahashi T and Alitalo K. (2002). J. Natl. Cancer Inst., 94, Busam KJ, Berwick M, Blessing K, Fandrey K, Kang S, 819–825. Karaoli T, Fine J, Cochran AJ, White WL, Rivers J. (1995). Hendrix MJ, Seftor RE, Seftor EA, Gruman LM, Lee LM, Am. J. Pathol., 147, 1049–1056. Nickoloff BJ, Miele L, Sheriff DD and Schatteman GC. Carmeliet P and Jain RK. (2000). Nature, 407, 249–257. (2002). Cancer Res., 62, 665–668. Carnochan P, Briggs JC, Westbury G and Davies AJ. (1991). Huang S, Mills L, Mian B, Tellez C, McCarty M, Yang XD, Br. J. Cancer, 64, 102–107. Gudas JM, Bar-Eli M. (2002). Am. J. Pathol., 161, Cascinelli N, Herlyn M, Schneeberger A, Kuwert C, Slominski 125–134. A, Armstrong C, Belli F, Lukiewcz S, Maurer D, Ansel J, Huang S, Xie K, Bucana CD, Ullrich SE and Bar-Eli M. Stingl G and Saida T. (2000). Exp. Dermatol., 9, (1996). Clin. Cancer Res., 2, 1969–1979. 439–451. Hubler Jr WR and Wolf Jr JE (1976). Cancer, 38, 187–192. Claffey KP, Brown LF, del Aguila LF, Tognazzi K, Yeo K-T, Hwu WJ (2000). Oncology, 14, 25–28. Manseau EJ and Dvorak HF. (1996). Cancer Res., 56, Hwu WJ, Krown SE, Panageas KS, Menell JH, Chapman PB, 172–181. Livingston PO, Williams LJ, Quinn CJ and Houghton AN. Dadras SS, Paul T, Bertoncini J, Brown LF, Mucikansky A, (2002). J. Clin. Oncol., 20, 2610–2615. Jackson DG, Ellwanger U, Garbe C, Mihm MC, Detmar M. Ilmonen S, Kariniemi AL, Vlaykova T, Muhonen T, Pyrhonen (2003). Am. J. Pathol., in press. S and Asko-Seljavaara S. (1999). Melanoma Res., 9, 273– Detmar M. (1996). J. Invest. Dermatol., 106, 207–208. 278. Detmar M and Hirakawa S. (2002). J. Exp. Med., 196, Karpanen T, Egeblad M, Karkkainen MJ, Kubo H, Yla- 713–718. Herttuala S, Jaattela M and Alitalo K. (2001). Cancer Res., Dotto GP, Moellmann G, Ghosh S, Edwards M and Halaban 61, 1786–1790. R. (1989). J. Cell. Biol., 109, 3115–3128. Kashani-Sabet M, Sagebiel RW, Ferreira CM, Nosrati M and Dreau D, Foster M, Hogg M, Swiggett J, Holder WD and Miller III JR (2002). J. Clin. Oncol., 20, 1826–1831. White RL. (2001). Oncol. Res., 12, 241–251. Kumar CC, Malkowski M, Yin Z, Tanghetti E, Yaremko B, Dvorak HF, Brown LF, Detmar M and Dvorak AM. (1995). Nechuta T, Varner J, Liu M, Smith EM, Neustadt B, Presta Am. J. Pathol., 146, 1029–1039. M, Armstrong L. (2001). Cancer Res., 61, 2232–2238. Eisen T, Boshoff C, Mak I, Sapunar F, Vaughan MM, Pyle L, Kusters B, Leenders WP, Wesseling P, Smits D, Verrijp K, Johnston SR, Ahern R, Smith IE, Gore ME (2000). Br. J. Ruiter DJ, Peters JP, van Der Kogel AJ and de Waal RM. Cancer, 82, 812–817. (2002). Cancer Res., 62, 341–345. Erhard H, Rietveld FJ, van Altena MC, Brocker EB, Ruiter Lacal PM, Failla CM, Pagani E, Odorisio T, Schietroma C, DJ and de Waal RM. (1997). Melanoma Res., 7(Suppl 2), Falcinelli S, Zambruno G and D’Atri S. (2000). J. Invest. S19–S26. Dermatol., 115, 1000–1007. Ferrara N. (1999). Curr. Top. Microbiol. Immunol., 237, 1–30. Lazar-Molnar E, Hegyesi H, Toth S and Falus A. (2000). Ferrara N and Alitalo K. (1999). Nat. Med., 5, 1359–1364. Cytokine, 12, 547–554. Folkman J. (1975). Ann. Intern. Med., 82, 96–100. Li Y, Wang MN, Li H, King KD, Bassi R, Sun H, Santiago A, Folkman J. (2002). Cancer Cell, 1, 113–115. Hooper AT, Bohlen P, Bicklin DJ (2002). J. Exp. Med., 195, Forsberg K, Valyi-Nagy I, Heldin CH, Herlyn M and 1575–1584. Westermark B. (1993). Proc. Natl. Acad. Sci. USA, 90, Lin P, Buxton JA, Acheson A, Radziejewski C, Maisonpierre 393–397. PC, Yancopoulos GD, Channon KM, Hale LP, Dewhirst Gitay-Goren H, Halaban R and Neufeld G. (1993). Biochem. MW, George SE and Peters KG. (1998). Proc. Natl. Acad. Biophys. Res. Commun., 190, 702–709. Sci. USA, 95, 8829–8834.

Oncogene Angiogenesis and melanoma M Streit and M Detmar 3179 Liu B, Earl HM, Baban D, Shoaibi M, Fabra A, Kerr DJ and Saeki H, Moore AM, Brown MJ and Hwang ST. (1999). J. Seymour LW. (1995). Biochem. Biophys. Res. Commun., 217, Immunol., 162, 2472–2475. 721–727. Salven P, Heikkila P and Joensuu H. (1997). Br. J. Cancer, 76, Lode HN, Moehler T, Xiang R, Jonczyk A, Gillies SD, 930–934. Cheresh DA and Reisfeld RA. (1999). Proc. Natl. Acad. Sci. Salven P, Lymboussaki A, Heikkila P, Jaaskela SH, Enholm USA, 96, 1591–1596. B, Aase K, von EG, Eriksson U, Alitalo K and Joensuu H. Luca M, Huang S, Gershenwald JE, Singh RK, Reich R and (1998). Am. J. Pathol., 153, 103–108. Bar-Eli M. (1997). Am. J. Pathol., 151, 1105–1113. Seftor RE, Seftor EA, Koshikawa N, Meltzer PS, Gardner Mandriota SJ, Jussila L, Jeltsch M, Compagni A, Baetens D, LM, Bilban M, Stetler-Stevenson WG, Quaranta V and Prevo R, Banerji S, Huarte J, Montesano R, Jackson DG, Hendrix MJ. (2001). Cancer Res., 61, 6322–6327. Orci L, Alitalo K, Christofori G and Pepper MS. (2001). Senger DR, Galli SJ, Dvorak AM, Perruzzi CA, Harvey VS EMBO J., 20, 672–682. and Dvorak HF. (1983). Science, 219, 983–985. Maniotis AJ, Folberg R, Hess A, Seftor EA, Gardner LM, Skobe M, Hamberg LM, Hawighorst T, Schirner M, Wolf GL, Pe’er J, Trent JM, Meltzer PS and Hendrix MJ. (1999). Am. Alitalo K and Detmar M. (2001a). Am. J. Pathol., 159, 893– J. Pathol., 155, 739–752. 903. Marcoval J, Moreno A, Graells J, Vidal A, Escriba JM, Garcia Skobe M, Hawighorst T, Jackson DG, Prevo R, Janes L, RM and Fabra A. (1997). J. Cutan. Pathol., 24, 212–218. Velasco P, Riccardi L, Alitalo K, Claffey K and Detmar M. Miao WM, Seng WL, Duquette M, Lawler P, Laus C and (2001b). Nat. Med., 7, 192–198. Lawler J. (2001). Cancer Res., 61, 7830–7839. Srivastava A, Laidler P, Davies RP, Horgan K and Hughes Mihm Jr MC, Clark Jr WH and Reed RJ. (1975). Semin. LE. (1988). Am. J. Pathol., 133, 419–423. Oncol., 2, 105–118. Srivastava A, Laidler P, Hughes LE, Woodcock J and Min HY, Doyle LV, Vitt CR, Zandonella CL, Stratton TJ, Shedden EJ. (1986). Eur. J. Cancer Clin. Oncol., 22, 1205– Shuman MA and Rosenberg S. (1996). Cancer Res., 56, 1209. 2428–2433. St Croix B, Rago C, Velculescu V, Traverso G, Romans KE, Morin MJ. (2000). Oncogene, 19, 6574–6583. Montgomery E, Lal A, Riggins GJ, Lengauer C, Vogelstein Muller A, Homey B, Soto H, Ge N, Cattron D, Buchanan B and Kinzler KW. (2000). Science, 289, 1197–1202. ME, McClanahan T, Murphy E, Yuan W and Wagner SN. Stacker SA, Baldwin ME and Achen MG. (2002). FASEB J., (2001). Nature, 410, 50–56. 16, 922–934. Naglich JG, Jure-Kunkel M, Gupta E, Fargnoli J, Henderson Stacker SA, Caesar C, Baldwin ME, Thornton GE, Williams AJ, Lewin AC, Talbott R, Baxter A, Bird J, Savopoulos R, RA, Prevo R, Jackson DG, Nishikawa S, Kubo H and Wills R, Kramer RA and Trail PA. (2001). Cancer Res., 61, Achen MG. (2001). Nat. Med., 7, 186–191. 8480–8485. Stenzinger W, Bruggen J, Macher E and Sorg C. (1983). Eur. J. Nesbit M, Nesbit HK, Bennett J, Andl T, Hsu MY, Dejesus E, Cancer Clin. Oncol., 19, 649–656. McBrian M, Gupta AR, Eck SL and Herlyn M. (1999). Straume O and Akslen LA. (2001). Am. J. Pathol., 159, 223– Oncogene, 18, 6469–6476. 235. Nurnberg W, Tobias D, Otto F, Henz BM and Schadendorf Straume O and Akslen LA. (2002). Am. J. Pathol., 160, 1009– D. (1999). J. Pathol., 189, 546–551. 1019. O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane Straume O, Salvesen HB and Akslen LA. (1999). Int. J. Oncol., WS, Flynn E, Birkhead JR, Olsen BR and Folkman J. 15, 595–599. (1997). Cell, 88, 277–285. Streit M, Stephen AE, Hawighorst T, Matsuda K, Lange- O’Reilly MS, Holmgren L, Shing Y, Chen C, Rosenthal RA, Asschenfeldt B, Brown LF, Vacanti JP and Detmar M. Moses M, Lane WS, Cao Y, Sage EH and Folkman J. (2002). Cancer Res., 62, 2004–2012. (1994). Cell, 79, 315–328. Ugurel S, Rappl G, Tilgen W and Reinhold U. (2001). J. Clin. Oliver G and Detmar M. (2002). Genes Dev., 16, 773–783. Oncol., 19, 577–583. Papoutsi M, Siemeister G, Weindel K, Tomarev SI, Kurz H, Vlaykova T, Laurila P, Muhonen T, Hahka-Kemppinen M, Schachtele C, Martiny BG, Christ B, Marme D and Wilting Jekunen A, Alitalo K and Pyrhonen S. (1999). Melanoma J. (2000). Histochem. Cell Biol., 114, 373–385. Res., 9, 59–68. Plum SM, Holaday JW, Ruiz A, Madsen JW, Fogler WE and Warren BA and Shubik P. (1966). Lab. Invest., 15, 464–478. Fortier AH. (2000). Vaccine, 19, 1294–1303. Weninger W, Uthman A, Pammer J, Pichler A, Ballaun C, Posey JA, Khazaeli MB, DelGrosso A, Saleh MN, Lin CY, Lang IM, Plettenberg A, Bankl HC, Sturzl M and Huse W and LoBuglio AF. (2001). Cancer Biother. Radio- Tschachler E. (1996). Lab. Invest., 75, 647–657. pharm., 16, 125–132. Wigle JT, Harvey N, Detmar M, Lagutina I, Grosveld G, Potgens AJ, Lubsen NH, van AM, Schoenmakers JG, Ruiter Gunn MD, Jackson DG and Oliver G. (2002). EMBO J., 21, DJ and de WR. (1995). Am. J. Pathol., 146, 197–209. 1505–1513. Prevo R, Banerji S, Ferguson DJ, Clasper S and Jackson DG. Wild R, Dhanabal M, Olson TA and Ramakrishnan S. (2000). (2001). J. Biol. Chem., 276, 19420–19430. Br. J. Cancer, 83, 1077–1083. Rak J, Yu JL, Kerbel RS and Coomber BL. (2002). Cancer Wiley HE, Gonzalez EB, Maki W, Wu MM and Hwang ST. Res., 62, 1931–1934. (2001). J. Natl. Cancer Inst., 93, 1638–1643. Reed JA, McNutt NS and Albino AP. (1994). Am. J. Pathol., Yayon A, Ma YS, Safran M, Klagsbrun M and Halaban R. 144, 329–336. (1997). Oncogene, 14, 2999–3009. Reiher FK, Volpert OV, Jimenez B, Crawford SE, Dinney CP, Yu JL, Rak JW, Carmeliet P, Nagy A, Kerbel RS and Henkin J, Haviv F, Bouck NP and Campbell SC. (2002). Int. Coomber BL. (2001). Am. J. Pathol., 158, 1325–1334. J. Cancer, 98, 682–689. Yu JL, Rak JW, Klement G and Kerbel RS. (2002). Cancer Rofstad EK and Halsor EF. (2000). Cancer Res., 60, 4932– Res., 62, 1838–1846. 4938. Yuan F, Chen Y, Dellian M, Safabakhsh N, Ferrara N and Rofstad EK and Halsor EF. (2002). Br. J. Cancer, 86, Jain RK. (1996). Proc. Natl. Acad. Sci. USA, 93, 14765– 301–308. 14770.

Oncogene