Meta-analytic magic, ivermectin, and socially responsible reporting.

Some clinicians prescribe ivermectin for COVID-19 despite a lack of support from any credible South African professional body. They argue that when faced by clinical urgency, weak signals of efficacy should trigger action if harm is unlikely. Several recent reviews found an apparent mortality benefit by including studies at high risk of bias and with active rather than placebo controls. If these studies are discounted, the pooled mortality effect is no longer statistically significant, and evidence of benefit is very weak. Relying on this evidence could cause clinical harm if used to justify vaccine hesitancy. Clinicians remain responsible for ensuring that guidance they follow is both legitimate and reliable. In the ivermectin debate, evidence-based medicine (EBM) principles have largely been ignored under the guise thatin a pandemic the 'rules are different', probably to the detriment of vulnerable patients and certainly to the detriment of the profession's image. Medical schools and professional interest groups are responsible for transforming EBM from a taught but seldom-used tool into a process of lifelong learning, promoting a consistent call for evidence-based and unconflicted debate integral to clinical practice.

Essential medicine selection during the COVID-19 pandemic: Enabling access in uncharted territory.

The COVID-19 pandemic requires urgent decisions regarding treatment policy in the face of rapidly evolving evidence. In response, the South African Essential Medicines List Committee established a subcommittee to systematically review and appraise emerging evidence, within very short timelines, in order to inform the National Department of Health COVID-19 treatment guidelines. To date, the subcommittee has reviewed 14 potential treatments, and made recommendations based on local context, feasibility, resource requirements and equity. Here we describe the rapid review and evidence-to-decision process, using remdesivir and dexamethasone as examples. Our experience is that conducting rapid reviews is a practical and efficient way to address medicine policy questions under pandemic conditions.

Data for indirect load case estimation of ice-induced moments from shaft line torque measurements.

During ice navigation, blade measurements of ice-induced moments on ship propellers, are challenged by the harsh operating environment. To overcome this problem, shaft line measurements are performed inboard, and the required propeller loads are subsequently estimated using a dynamic model and the solution of an inverse problem. The inverse problem is mathematically ill-posed and requires the determination of the ice-induced moment on the propeller blades from shaft line measurements. Full-scale torsional response data is presented as calculated from indirect strain measurements on the shaft line of a polar supply and research vessel. The vessel operated on a 68-day voyage between Cape Town and Antarctica and spent almost 11 days in sea ice with observed concentrations above 90% and a maximum thickness of 3 m. Data for five ice-induced load cases are presented, including the shaft torque from indirect measurements and the estimated ice-induced moment, which is obtained by solving an ill-posed inverse problem. The ice-induced moments on the propeller are obtained by approximating the drive-train as a viscously damped, elastic lumped mass model. The ice-induced moment is then determined through existing approaches to solving the ill-conditioned inverse problem. The lumped mass model is presented along with algorithms to solve the inverse problem, including truncated singular value decomposition, truncated generalized singular value decomposition and Tikhonov׳s method. The resulting time series data for the inversely calculated ice-induced moments is published to provide industry with load cases for ice-going propulsion design.

Clinician Compliance with Laboratory Monitoring and Prescribing Guidelines in HIV-1-Infected Patients Receiving Tenofovir

BACKGROUND:Tenofovir is part of the preferred first-line regimen for HIV-infected patients in South Africa (SA); but is associated with kidney toxicity. SA antiretroviral therapy (ART) guidelines recommend creatinine monitoring at baseline (ART start) and at 3; 6 and 12 months; and substituting tenofovir with zidovudine; stavudine or abacavir should creatinine clearance (CrCl) decrease to etlt;50 mL/min. OBJECTIVE:To assess clinician compliance with tenofovir monitoring and prescribing guidelines.METHODS:We described the proportion of adult patients on tenofovir-based first-line ART who were screened for baseline renal impairment; were monitored according to the SA antiretroviral treatment guidelines; and were switched from tenofovir if renal function declined.RESULTS:We included 13 168 patients who started ART from 2010 to 2012. Creatinine concentrations were recorded in 11 712 (88.9%) patients on tenofovir at baseline; 9 135/11 657 (78.4%) at 3 months; 5 426/10 554 (51.4%) at 6 months; and 5 949/ 8 421 (70.6%) at 12 months. At baseline; 227 (1.9%) started tenofovir despite a CrCl etlt;50 mL/min. While on tenofovir; 525 patients had at least one CrCl of etlt;50 mL/min. Of 382 patients with =3 months' follow-up after a CrCl etlt;50 mL/min; 114 (29.8%) stopped tenofovir within 3 months. Clinicians were more likely to stop tenofovir in patients with lower CrCl and CD4 count. Of 226 patients who continued to receive tenofovir and had further CrCls available; 156 (69.0%) had a CrCl =50 mL/min at their next visit.CONCLUSIONS:Creatinine monitoring is feasible where access to laboratory services is good. Kidney function recovered in most patients who continued to receive tenofovir despite a CrCl etlt;50 mL/min. Further research is needed to determine how best to monitor renal function with tenofovir in resource-limited settings

Plasminogen activators are involved in angiostatin generation in vivo in benign and malignant ovarian tumor cyst fluids.

In many tumor types, angiogenesis is the net result of pro- and anti-angiogenic mediators and correlated with metabolic activity, growth, and degree of malignancy. One of the first discovered anti-angiogenic compounds is angiostatin, a proteolytic fragment of plasminogen. The requirements for in vivo angiostatin generation have not yet been determined. We investigated the levels of plasminogen and angiostatin by western blotting and of components of the plasminogen activator complex by ELISA in cyst fluid derived from benign and malignant ovarian tumors. Fluid samples from functional ovarian follicles, dermoid cysts and endometriotic lesions were evaluated separately. When no or minimal amounts of plasminogen were present in the cyst fluids, angiostatin was generally absent as well, irrespective of plasminogen activator concentrations. When plasminogen was present, the degree of conversion of plasminogen to angiostatin was significantly correlated with the level of uPA, and, to a lesser extent, to the tPA level. However, angiostatin was also found in a number of cyst fluid samples with minimal or no plasminogen activators, suggesting the involvement of other angiostatin generating proteases in these samples. Conversely, no angiostatin was observed in a number of cyst fluid samples containing both plasminogen and plasminogen activators. The presence of an inhibitor of the enzymatic activity of uPA and/or tPA, like PAI-1, may explain this finding. Our data show that plasminogen activators are clearly involved in in vivo angiostatin formation in ovarian cysts. Most likely, however, other proteases, as well as inhibitors of plasminogen activators, are involved as well.

Limited expression of heparan sulphate proteoglycans associated with Aß deposits in the APPswe/PS1dE9 mouse model for Alzheimer's disease.

AIMS: Alzheimer's disease (AD) is characterized by deposition of the amyloid beta (Aß) peptide in brain parenchyma and vasculature. Several proteins co-deposit with Aß, including heparan sulphate proteoglycans (HSPG). HSPG have been suggested to contribute to Aß aggregation and deposition, and may influence plaque formation and persistence by stimulating Aß fibrillization and by protecting Aß against degradation. Mouse models for AD, expressing the human amyloid precursor protein (APP), produce Aß deposits similar to humans. These models may be used to study disease pathology and to develop new therapeutic interventions. We aimed to investigate whether co-deposition of HSPG in AD brains can be replicated in the APPswe/PS1dE9 mouse model for AD and if a temporal association of HSPG with Aß exists. METHODS: We studied the co-deposition of several HSPG and of the glycosaminoglycan side chains of HSPG in the APPswe/PS1dE9 model at different ages by immunohistochemistry. RESULTS: We found that, although APPswe/PS1dE9 mice did develop severe Aß pathology with age, co-deposition of HS glycosaminoglycan chains and the various HSPG (agrin, perlecan and glypican-1) was scarce (<10-30% of the Aß deposits were stained). CONCLUSIONS: Our data suggest that the molecular composition of Aß deposits in the APPswe/PS1dE9 mouse, with respect to the several HSPG investigated in this study, does not accurately reflect the human situation. The near absence of HSPG in Aß deposits in this transgenic mouse model may, in turn, hinder the translation of preclinical intervention studies from mice to men.

Micronodular transformation as a novel mechanism of VEGF-A-induced metastasis.

How and why tumors metastasize is still a matter of debate. The assumption is that mutations render tumor cells with a metastatic phenotype, enabling entrance in and transport through lymph or blood vessels. Distant outgrowth is thought to occur only in a suitable microenvironment (the seed and soil hypothesis). However, the anatomical location of most metastases in cancer patients suggests entrapment of tumor cells in the first microcapillary bed that is encountered. We here investigated how vascular endothelial growth factor-A (VEGF-A) attributes to the metastatic process. We describe here that VEGF-A enhances spontaneous metastasis by inducing intravasation of heterogeneous tumor cell clusters, surrounded by vessel wall elements, via an invasion-independent mechanism. These tumor clusters generate metastatic tissue embolisms in pulmonary arteries. Treatment of tumor-bearing mice with the antiangiogenic compound ZD6474 prevented the development of this metastatic phenotype. This work shows that tumors with high constitutive VEGF-A expression metastasize via the formation of tumor emboli and provides an alternative rationale for anti-VEGF-A therapy, namely to inhibit metastasis formation.

Specific association of small heat shock proteins with the pathological hallmarks of Alzheimer's disease brains.

The small heat shock protein family (sHsp) comprises molecular chaperones able to interact with incorrectly folded proteins. Alzheimer's disease (AD) is characterized by pathological lesions such as senile plaques (SPs), cerebral amyloid angiopathy (CAA) and neurofibrillary tangles (NFTs), predominantly consisting of the incorrectly folded proteins amyloid-beta (Abeta) and tau respectively. The aim of this study was to investigate the association of the chaperones Hsp20, HspB2, alphaB-crystallin and Hsp27 with the pathological lesions of AD brains. For this purpose, a panel of well-characterized antibodies directed against these sHsps was used in immunohistochemistry and immunoblotting. We observed extracellular expression of Hsp20, Hsp27 and HspB2 in classic SPs, and Hsp20 expression in diffuse SPs. In addition, extracellular expression of HspB2 was observed in CAA. Both Hsp27 and alphaB-crystallin were also observed in astrocytes associated with both SPs and CAA. Furthermore, none of the sHsps were observed in NFTs in AD brains. We conclude that specific sHsp species may be involved in the pathogenesis of either SPs or CAA in AD.

Differential gene expression in human brain pericytes induced by amyloid-beta protein.

Cerebral amyloid angiopathy is one of the characteristics of Alzheimer's disease (AD) and this accumulation of fibrillar amyloid-beta (Alphabeta) in the vascular wall is accompanied by marked vascular damage. In vitro, Abeta1-40 carrying the 'Dutch' mutation (DAbeta1-40) induces degeneration of cultured human brain pericytes (HBP). To identify possible intracellular mediators of Abeta-induced cell death, a comparative cDNA expression array was performed to detect differential gene expression of Abeta-treated vs. untreated HBP. Messenger RNA expression of cyclin D1, integrin beta4, defender against cell death-1, neuroleukin, thymosin beta10, and integrin alpha5 were increased in DAbeta1-40-treated HBP, whereas insulin-like growth factor binding protein-2 mRNA expression was decreased. Corresponding protein expression was investigated in AD and control brains to explore a potential role for these proteins in pathological lesions of the AD brain. Cyclin D1 expression was increased in cerebral amyloid angiopathy and cells in a perivascular position, suggesting that the cell cycle may be disturbed during Abeta-mediated degeneration of cerebrovascular cells. Moreover, cyclin D1 expression, but also that of integrin beta4, defender against cell death-1, neuroleukin and thymosin beta10 was found in a subset of senile plaques, suggesting a role for these proteins in the pathogenesis of senile plaques.

Lack of lymphangiogenesis despite coexpression of VEGF-C and its receptor Flt-4 in uveal melanoma.

PURPOSE: Because lymphatic vessels are absent from the normal eye and because uveal melanomas are presumed to spread by a hematogenous route in the absence of tumor exposure to conjunctival lymphatics, this study was undertaken to investigate the presence of lymphatic vessels in primary uveal melanomas. METHODS: The presence of lymphatics in 2 control eyes and in 33 primary uveal, 10 primary cutaneous, and 3 metastatic cutaneous melanomas was evaluated by using a double-immunostaining protocol that differentially highlights blood and lymphatic vasculature. In addition, 14 uveal melanomas were immunostained for the lymphatic growth factor vascular endothelial growth factor (VEGF)-C (with anti-VEGF-C polyclonal antibodies [pAbs]), its receptors Flt-4 (with monoclonal antibody [mAb] 9D9) and KDR (with anti-KDR mAb [Clone KDR-2]), and the hemangiogenic factor VEGF-A (with anti-VEGF pAbs). RESULTS: Lymphatics were not detected in normal eyes or in uveal melanoma. As a consequence, signs of lymphangiogenesis were not present. There was coexpression of VEGF-C with Flt-4 and KDR in 6 (43%) of the 14 melanomas. Staining for VEGF-A was completely negative in 25 uveal melanomas analyzed. CONCLUSIONS: The strictly hematogenous metastasis of primary uveal melanomas is explained by the absence of lymphatics in and around the tumor. The current data suggest that, in the presence of endothelial Flt-4, VEGF-C expression is not sufficient to induce lymphangiogenesis from preexisting blood vessels in human cancer.

The mitochondrial toxin 3-nitropropionic acid induces differential expression patterns of apoptosis-related markers in rat striatum.

The mitochondrial toxin 3-nitropropionic acid (3-NP) causes selective striatal lesions in rats and serves as an experimental model for the neurodegenerative disorder Huntington's disease (HD). Apoptotic cell death has been implicated for the neuronal degeneration that occurs in HD brains. The present study was designed to investigate whether the 3-NP-induced cell death in rats involves apoptosis and an altered expression of Bcl-2 family proteins. Systemic administration of 3-NP via subcutaneous Alzet pumps resulted in lesions of variable severity with neuronal loss and gliosis in the striatum. Using the terminal transferase-mediated biotinylated-UTP nick end-labelling (TUNEL) of DNA, TUNEL-positive cells exhibiting typical apoptotic morphology were detected only in the striatum of rats with a severe lesion. Furthermore, the neuronal expression of the pro-apoptotic protein Bax was strongly increased in the core of the severe lesion. Expression of the anti-apoptotic marker Bcl-2 was unchanged in this location, but was enhanced in the margins of the lesions. A moderately increased expression of both Bax and Bcl-2 was observed in dark neurones in the mild lesion and in the subtle lesion. The presence of nuclear DNA fragmentation, strong granular Bax expression and an increased Bax/Bcl-2 ratio in the centre of severe lesions suggests the occurrence of apoptotic cell death following 3-NP administration. In contrast, the dark compromised neurones observed in 3-NP-treated animals revealed an equally enhanced expression of both Bax and Bcl-2, but lacked TUNEL-labelling, and are therefore not apoptotic.

The pattern of metastasis of human melanoma to the central nervous system is not influenced by integrin alpha(v)beta(3) expression.

We investigated the effect of integrin alpha(v)beta(3) expression on the metastatic pattern of human melanoma cells in the central nervous system (CNS). For this purpose, we developed a hematogenous CNS melanoma metastasis model in nude mice using a modified internal carotid artery infusion technique. This protocol revealed 2 different patterns of CNS metastasis. The integrin alpha(v)beta(3)-expressing melanoma lines Mel57 and Zkr nearly exclusively produced metastases in the brain parenchyma, whereas cells of the BLM and MV3 lines, devoid of integrin alpha(v)beta(3) expression, preferentially metastasized to dura mater and leptomeninges. Treatment with hyaluronidase to obtain single BLM cell suspensions did not influence the metastatic pattern, indicating that this was not simply the result of entrapment of tumor cell aggregates in large-sized leptomeningeal vessels. The role of integrin alpha(v)beta(3) expression in the process of metastasis was tested by transfection of BLM, but did not lead to an altered pattern of metastasis. We did observe, however, slower growth of the transfected tumors, although the in vitro growth rate was unaltered, indicating a reduction in tumorigenicity. We conclude from our findings that CNS metastasis of melanoma cells in the mouse xenograft model occurs in at least 2 different but very reproducible patterns. Although it is predicted that adhesion of tumor cells to endothelial cells plays a role in this phenomenon, tumor cell integrin alpha(v)beta(3) expression per se does not explain the difference in metastatic behavior in the CNS. We assume that other, as yet unknown factors, must be involved.

Lymphangiogenesis in malignant tumours: Does it occur?

The development of a vascular bed is essential for solid tumour growth and metastasis. In many tumours, mean vascular density can be related to the rate of metastasis and, therefore, to prognosis. In other tumour types, such as cutaneous melanoma and head-and-neck squamous cell carcinoma, this relation is absent. Until now, the reason for this has been unclear, but since these particular tumour types are also known for their propensity to spread via the lymphatic system, it may be speculated that the presence of a pre-existing lymphatic bed and the formation of new lymphatics (lymphangiogenesis) are important factors. Growth factors involved in lymphangiogenesis during embryogenesis have been recently identified and these are also expressed in many tumour types, but the existence of tumour-induced lymphangiogenesis has not so far been reported. Partly, this could be due to the lack of reliable endothelial markers, thereby hampering a consistent evaluation of lymphatic vasculature. This editorial discusses the role of the lymphatic bed in mediating the metastasis of solid tumours, summarizes known methods to detect lymphatics, and proposes a hypothetical mechanism of tumour-induced lymphangiogenesis.

In vivo activities of mutants of vascular endothelial growth factor (VEGF) with differential in vitro activities.

VEGF mutants in which Cys51 or Cys60 are converted into a serine are poor inducers of proliferation in human umbilical vein endothelial cells, but they have wild-type activity in the Miles vascular permeability assay. To assess the contribution of proliferation vs. other VEGF activities such as vascular permeability, to tumor angiogenesis and growth, C127I cells, transfected with BPV-based expression plasmids carrying wild-type or mutated VEGF cDNAs, were injected subcutaneously in BALB/c nu/nu mice. From C127I cells expressing wtVEGF(165), intensely vascularized and invasive tumors developed within 2 to 3 weeks. From cells expressing VEGF-Cys51Ser or VEGF-Cys60Ser, tumors developed only after 2 to 3 months, comparable to the time of development of control tumors, i.e., tumors from cells transfected with empty vector. Despite the late take, the VEGF-Cys51Ser and VEGF-Cys60Ser tumors developed an extensive vascular bed with an architecture comparable to that of recombinant wtVEGF-producing tumors whereas control tumors had a considerably lower vascular density. No metastases were detected in mice carrying either wtVEGF or mutant VEGF expressing tumors. Thus, because proliferation-defective VEGF-mutants cannot induce angiogenesis, we conclude that the proliferation-inducing effect of VEGF is crucial for tumor angiogenesis and growth. The hypervasculature in the tumors expressing these VEGF-mutants suggests, however, that other VEGF-activities, such as the induction of vascular permeability, strongly affects vascular density and vascular structure. Furthermore, neither overexpression of VEGF or a high vascular density or hyperpermeability of tumor vasculature is necessarily followed by metastasis.

Tumour metastasis: is tissue an issue?

Single tumour cells, or multicellular aggregates, escape into blood and lymphatic vessels from a primary solid tumour as it progresses. However, the ability of such cells to develop into metastatic outgrowths is very limited, as shown by the poor prognostic power of the presence of circulating tumour cells in cancer patients. An explanation for this low efficiency may be a temporary absence of a suitable microenvironment once the tumour cells escape from their original tissue compartment. On the basis of histopathological observations, experimental studies, and the generally accepted poor prognosis of histopathologically confirmed intravascular tumour location, we propose that the structural and functional organisation of intravascular tumour cells as a tissue has a key role in providing the optimum microenvironment for sustained malignant dissemination. Such a tissue may be a fixed or a mobile intravascular microsatellite, or an intravascular micrometastasis, which locally develops into an overt in-transit metastasis.

Heparan sulfate proteoglycan expression in cerebrovascular amyloid beta deposits in Alzheimer's disease and hereditary cerebral hemorrhage with amyloidosis (Dutch) brains.

Cerebrovascular deposition of amyloid beta protein (A beta) is a characteristic lesion of Alzheimer's disease (AD) and hereditary cerebral hemorrhage with amyloidosis of the Dutch type (HCHWA-D). Besides A beta, several other proteins and proteoglycans accumulate in cerebral amyloid angiopathy (CAA). We have now analyzed the expression of the heparan sulfate proteoglycan (HSPG) subtypes agrin, perlecan, glypican-1, syndecans 1-3 and HS glycosaminoglycan (GAG) side chains in CAA in brains of patients with AD and HCHWA-D. Hereto, specific well-characterized antibodies directed against the core protein of these HSPGs and against the GAG side chains were used for immunostaining. Glypican-1 was abundantly expressed in CAA both in AD and HCHWA-D brains, whereas perlecan and syndecans-1 and -3 were absent in both. Colocalization of agrin with vascular A beta was clearly observed in CAA in HCHWA-D brains, but only in a minority of the AD cases. Conversely, syndecan-2 was frequently associated with vascular A beta in AD, but did not colocalize with vascular A beta deposits in HCHWA-D. The three different syndecans, agrin, glypican-1 and HS GAG, but not perlecan, were associated with the majority of senile plaques (SPs) in all brains. Our results suggest a role for agrin in the formation of SPs and of CAA in HCHWA-D, but not in the pathogenesis of CAA in AD. Both syndecan-2 and glypican, but not perlecan, may be involved in the formation of CAA. We conclude that specific HSPG species may be involved in the pathogenesis of CAA in both AD and HCHWA-D, and that the pathogenesis of CAA and SPs may differ with regard to the involvement of HSPG species.

Angiostatin generation by human tumor cell lines: involvement of plasminogen activators.

Angiostatin is a tumor-derived angiogenesis inhibitor consisting of an internal fragment of plasminogen. Little is known about the production of angiostatin by human tumors. In this study, we examined the in vitro angiostatin-generating capacities of a panel of human tumor cell lines (total n = 75) and the proteolytic molecule(s) involved. Angiostatin formation was determined by assessing the level of plasminogen digestion in conditioned medium by Western-blot analysis. We found that the capacity to produce angiostatin is a common feature of many cell lines, depending on the tumor type. All 6 bladder-carcinoma and 6 out of 7 prostate-carcinoma cell lines showed intermediate to potent angiostatin-generating activity. In contrast, only 2 out of 7 colon-carcinoma and 2 out of 9 renal-cell carcinoma cell lines were able to generate angiostatin at intermediate levels. Out of 25 melanoma cell lines, only one line failed to generate angiostatin. In the other cell-line groups (cervix, breast and ovary), angiostatin formation varied. Remarkably, angiostatin bands were not of equal size in all plasminogen digests. Since reported data have indicated that plasminogen activators (uPA and tPA) were able to excise the angiostatin fragment from the plasminogen parent molecule via plasmin generation, we determined levels of uPA and tPA and PAI-1 antigen in the conditioned media, and correlated the results with angiostatin-generating capacity. Whereas prostate- and bladder-carcinoma lines capable of generating high levels of angiostatin showed high uPA levels, angiostatin generation in melanoma cell lines was correlated with tPA levels. Generally, angiostatin non-producers did not express uPA or tPA. In 6 out of 75 cell lines, however, we found angiostatin generation combined with low or absent levels of plasminogen activator, suggesting the involvement of alternative proteolytic pathways in the generation of angiostatin.

Angiogenic balance in human melanoma: expression of VEGF, bFGF, IL-8, PDGF and angiostatin in relation to vascular density of xenografts in vivo.

Tumor angiogenesis, a major requirement for tumor outgrowth and metastasis formation, is regulated by pro- and anti-angiogenic factors. We have studied the expression of a panel of angiogenic factors, and of the angiogenesis inhibitor angiostatin, in a panel of human melanoma cell lines giving rise to xenografts with different vascular densities. Angiogenic-factor expression was analyzed in vitro (cell lines) and in vivo (xenografts), both at mRNA (RT-PCR and Northern blot) and at protein level (ELISA and Western blot). In vitro angiostatin generation was assessed by Western-blot analysis. Expression of bFGF and VEGF was clearly correlated with a high degree of vascularization, confirming the importance of these factors for tumor angiogenesis. In addition, there was exclusive or elevated in vitro expression of angiogenic factors IL-8, PDGF-AB, and, to a lesser extent, midkine in cell lines that formed highly vascularized tumors. A similar angiogenic-factor-expression pattern was found in the corresponding xenografts, with the exception of VEGF. In most cell lines, this factor had low expression in vitro which was strongly enhanced in vivo. Although all 8 melanoma cell lines were able to excise the angiostatin fragment from the plasminogen parent molecule in vitro, cell lines BLM and M14 showed the most potent angiostatin generation. In vitro angiostatin generation by cell lysates prepared from melanoma xenografts was comparable in all xenograft types. Thus, in our model system we found no correlation between angiostatin generation and vascular density. Our study has limited the number of pro-angiogenic factors that may be involved in melanoma angiogenesis, and provides evidence for the notion that regulation of tumor angiogenesis is dependent on multiple factors. Inhibition of angiogenesis for therapeutic purposes, therefore, should preferably not concentrate on a single factor.