Anti-angiogenic therapy for uveal melanoma—more haste, less speed

Vascularisation is critical for the support of substantial tumour growth. For a wide range of tumours, a complex microvasculature accompanies the transition from hyperplasia to neoplasia, a progression from low to high grade classification and enhanced metastatic capacity. The pioneering work of Folkman and Warren and Shubik outlined the critical importance of “tumour angiogenesis” to an initially sceptical field as far back as the late 1960s.^1,2 Folkman's struggle to have this hypothesis supported has been recently portrayed in a fine biography by Robert Cooke.³ Thirty years on, it is now accepted that most tumours require a complex microvasculature in order to grow beyond approximately 1–2 mm in diameter. The belief that their essential, and potentially labile, vascular supply represents a chink in the tumour armour has precipitated an ever growing raft of research. Indeed, tumour angiogenesis is a key concept that supports investigations in a range of clinically relevant research areas such as the diagnosis and prognosis of cancers, discovery and molecular characterisation of new angiogenic factors, the identification and therapeutic potential of endogenous anti-angiogenic agents, and characterisation of tumour specific vascular markers.

Pathogenic angiogenesis is a well known phenomenon to ophthalmologists and vision scientists since it is the underlying basis of many important ocular diseases such as retinopathy of prematurity, proliferative diabetic retinopathy, and the wet form of age related macular degeneration. Not surprisingly, angiogenesis is also thought to be a central event in growth of ocular neoplasms, none more so than uveal melanomas, where a high tumour vascularity index carries increased risk of metastasis and poor prognostic outcome. They are also known to precipitate secondary neovascularisation of the retina and iris. However, there are significant and extensive gaps in our knowledge of uveal melanomas and their secondary effects. We are still seeking answers to key questions such as what exactly is the cellular and molecular basis of angiogenic growth in these tumours, how does this affect remote, non-involved ocular tissues such as the retina, and what target(s) represent the best therapeutic options for future development?

Angiogenesis is . . . thought to be a central event in growth of ocular neoplasms . . . where a high tumour vascularity index carries increased risk of metastasis and poor prognostic outcome

In this issue of the BJO, Boyd et al (pp 440 and 448) present data that confirm an important role for vascular endothelial growth factor-A (VEGF-A) and, to a lesser extent, basic fibroblast growth factor (bFGF) in angiogenic growth in and around uveal melanomas. Using a range of approaches, they assayed these factors in ocular fluids and postmortem specimens from patients, grew co-cultures of tumour cells and endothelium in vitro, and concluded that anti-angiogenic therapy may be a worthwhile approach for treatment of eyes harbouring uveal melanomas. In many of these tumours VEGF-A was expressed at high levels, and this was especially true in patients who developed secondary retinal and iris neovascularisation following ionising radiation treatment of their tumours. VEGF-A expression was shown to be particularly high in the ocular fluids of this subset of patients and Boyd et al conclude that this phenomenon makes the case for anti-VEGF-A treatment as an adjunctive therapy in the treatment of uveal melanoma and secondary neovascularisation. While this may be so, the authors acknowledge that the situation is still extremely complex and interpatient variability remains surprisingly high. Furthermore, it should be considered that increased levels of VEGF following irradiation of the tumour could represent a classic response to tissue hypoxia following enhanced destruction of vascular endothelial cells that, in the case of melanoma, may be more radiosensitive than the tumour they serve.

While anti-angiogenic therapies continue to be developed at an ever increasing pace, it is perhaps worth considering the pros and cons of such approaches. As a case in point, uveal melanomas carry a serious risk of metastasis and enucleation remains a necessary, if last resort, option. Therefore, there are real pressures to take radical steps that may reduce growth and metastatic potential of these aggressive tumours and prevent secondary neovascularisation. Agents such as anti-VEGF antibodies, VEGF receptor or integrin antagonists, VEGF aptamers, matrix metalloproteinase (MMP) inhibitors, antisense oligonucleotides to growth factor mRNAs, and established drugs such as thalidomide are already in various stages of clinical trial and may offer therapeutic benefits. Given the proliferation of these anti-angiogenics and the prolonged nature of many trials we will soon discover the validity of such an approach. However, it should be considered that growth factors such as VEGF-A represent single members of highly complex and inter-related families, many of which can also contribute to ocular angiogenesis. Even more importantly, many of these peptides are proved survival factors, especially in the retinal microvasculature.^4,5 and their effective depletion in a delicately balanced system could have serious long term and unwanted vasodegenerative effects in non-target tissue.

Another potentially negative consequence of an anti-angiogenic strategy for uveal melanomas and/or secondary neovascularisation, especially as an adjunct to radiation therapy, may be the inadvertent promotion of a more aggressive tumour phenotype. It has been demonstrated in various malignant tumours that induction of hypoxia, by restricting vascular supply, may actually promote selection of hypoxia resistant tumour cells with decreased apoptotic potential. An initial reduction in tumour size may be followed by a more malignant counterpart (for a review of this concept see Hockel and Vaupel⁶). It is clear, therefore, that during treatment of the neovascular consequences of uveal melanomas with anti-angiogenics, their effects on tumour growth and stability will need to be carefully monitored. In addition, recent evidence also points towards anti-angiogenic therapy (in particular through blocking VEGF-A bioactivity) that serves to “normalise” tumour vasculature, which in turn promotes effective delivery of oxygen that is essential for efficacious radiotherapy.⁷ In short, it may be important to make use of anti-angiogenic approaches in a carefully planned and titrated manner that can enhance therapy of solid tumours.

Uveal melanomas represent a serious ocular problem and much research is needed in order to advance the established ophthalmic approaches to these tumours. Using fundamental laboratory and clinic based research, the angiogenic characteristics of the vasculature in uveal melanomas should be carefully determined. This is important, since residence in a highly oxygenated microenvironment such as the choroid may make uveal melanomas distinct from cutaneous counterparts. Also, recent research, indicates that the vasculature of many tumours exists in a dynamic equilibrium of mature and immature vessels and that highly malignant varieties (such as glioblastomas) have a very high ratio of immature vessels.⁵ This subpopulation of vessels is often the specific target of anti-angiogenic agents. This being the case, new tumour specific and tumour vasculature specific targets should be identified that can ultimately allow precisely targeted treatments, that will not significantly upset the delicate angiogenic balance that exists in many ocular tissues. Following on from this, novel delivery systems should be developed, which will allow tumours lodged in the posterior uvea to have efficacious agents delivered at appropriate concentrations while minimising risk to the non-involved retina or choroid. Efficient transcleral delivery is being researched and, if possible, may provide a approach that is particularly relevant to uveal melanomas.