Antivascular therapy: a new approach to cancer treatment
BMJ 1999; 318 doi: https://doi.org/10.1136/bmj.318.7187.853 (Published 27 March 1999) Cite this as: BMJ 1999;318:853
- A J Hayes, surgical registrar (hayesa{at}gunet.georgetown.edu),
- L Y Li, assistant professor of biochemistry,
- M E Lippman, professor of medicine and pharmacology
- Lombardi Cancer Center, Georgetown University Medical Center, 3970 Reservoir Road, Washington DC, 20007, USA
- Correspondence to: Dr Hayes
The understanding that the growth of tumours is dependent on angiogenesis has led to the development of new approaches to treatment and new agents directed at tumour vasculature. These have yielded striking successes in experimental models, which if translated into the clinical setting will have a substantial effect on patient survival. Such new approaches are vital because, although great strides have occurred in the treatment of certain cancers, the overall standardised mortality from most solid tumours has altered little over the past two decades.1
This article considers the process of tumour angiogenesis and discusses the potential of angiogenic inhibitors as therapeutic agents.
Role of angiogenesis in growth of tumours
The vascularity of tumours has been noted for many years.2 Alguire noted that vascularisation was instigated by the developing tumour: “An outstanding characteristic of the growing tumour is its capacity to elicit the production of a new capillary endothelium from the host.”3 Tannock elegantly showed that the rate of division of tumour cells decreased in proportion to their distance from the supplying blood vessel and related this to diminishing oxygen supply.4 Moreover, he showed that the overall rate of growth was dictated not by proliferation of tumour cells but by the lower rate of proliferation of endothelial cells, concluding that the supply of oxygen and nutrients to the tumour limited its growth.
Predicted developments
Research will clarify the mechanisms by which endogenous inhibitors of angiogenesis prevent tumour growth
Strategies will be developed for large scale production of antivascular drugs for clinical use
New treatment regimens will be developed to modify the balance of positive and negative angiogenic proteins in tumours
Extensive clinical evaluation of new antivascular treatments alongside traditional treatments will define their anticancer potential more clearly
New trials and treatments will focus on inducing long term remission
Tumour vascularisation is a vital process for the progression of a neoplasm from a small, localised tumour to an enlarging tumour with the ability to metastasise (figure). 5 6 Anti-angiogenesis as a therapeutic concept was developed in the early 1970s based on observations that tumours that did not vascularise failed to grow beyond a few millimetres in diameter.7 By comparing the growth of transplanted tumours in the avascular aqueous humour of a rabbit eye with those in the vascular iris, Folkman could show distinct avascular and vascular phases of tumour growth. The start of the vascular phase of growth coincided with tumours growing beyond 2-3 mm3 and a 20-fold increase in the rate of tumour growth. Tumours in the aqueous humour were prevented from entering the vascular phase and remained dormant.8 He concluded that vascularisation was essential to tumour growth and inferred that preventing this process was a viable therapeutic approach.
Induction of angiogenesis by tumours
Adult endothelium is essentially quiescent, but in response to physiological or pathological stimuli (such as proliferating endometrium, injury, tumour growth, or diabetic retinopathy) the endothelium can alter to a proliferating and organising population of cells. Physiological angiogenesis can also be rapidly curtailed, indicating that the process is held in check physiologically and yet can be activated in response to the appropriate stimuli, somewhat analogous to the clotting cascade. Box 1 lists some of the regulatory proteins identified.
A tumour induces this proliferative vascular response from host vessels by altering the balance of positive and negative regulators locally. This “angiogenic switch” is necessary for tumour growth and may be rate limiting.9 Convincing evidence exists that tumours undergo a switch to an angiogenic character as they progress. In cervical carcinoma the development of vascularity can be associated with progression from a non-invasive premalignant stage to invasive carcinoma.10 The density of microvessels in tumours is a powerful independent prognostic indicator of distant metastasis and survival, suggesting that tumour vascularisation correlates with growth and metastatic potential.11
Box 1 Proteins that may regulate angiogenesis
Antivascular treatment
There are four key approaches to antivascular treatment (box 2). All depend on targeting endothelial cells rather than tumour cells for drug action, and destruction of the tumour cells is secondary. A theoretical advantage of these approaches is that endothelial cells are not transformed and are unlikely to acquire mutations resulting in drug resistance. Furthermore, treatment directed at endothelial cells is applicable to all solid tumours, irrespective of the origin of the tumour cells. Also, endothelial cells are uniquely exposed to bloodborne agents, circumventing the problem of delivering drugs to the centr of a tumour, which is a major hurdle in conventional treatment.
Box 2 Approaches to antivascular treatment
Neutralising angiogenic promoters
Interfere with the positive effect of angiogenic factors produced by a tumour
Examples include
Antibodies to vascular endothelial growth factor
mViral delivery of dominant negative receptors to vascular endothelial growth factor
Prevent release and activation of fibroblast growth factor 2
Endogenous angiogenic inhibitors
Use endogenous inhibitory proteins to counter the angiogenic stimulus produced by tumours
Examples include
Supply angiogenic inhibitors directly—such as angiostatin, endostatin
Gene transfer of DNA to angiogenesis inhibitors—
angiostatin, platelet factor 4
Endothelial cell targets
Use specific markers for tumour endothelial cells to direct a toxin or antibody to tumour vasculature to cause tumour infarction
Targets include
Integrin αvβ3
Vascular endothelial growth factor-receptor complexes
Endoglin
Synthetic angiogenic inhibitors
Inhibit tumour angiogenesis with drugs that specifically prevent endothelial cell division
These drugs can act synergistically with conventional cytotoxic treatment and with “hypoxia activated” cytotoxic drugs
Neutralising angiogenic promoters
One of the favoured approaches is to interfere with the balance of angiogenic proteins produced by a tumour—turning off the “switch.” This has been made possible by an understanding of the mediators of angiogenesis in normal and pathological settings. Current evidence implicates vascular endothelial growth factor (VEGF) and the family of fibroblast growth factors (FGF 1-12) as critical regulators of physiological angiogenesis. Angiopoietin 1 and 2 have been implicated as the factors controlling the recruitment of supporting cells to the developing tubule. 12 13
Both vascular endothelial growth factor and fibroblast growth factors 1 and 2 have been implicated as important promoters of tumour angiogenesis. Vascular endothelial growth factor, a mitogen and permeability factor specific for endothelial cells, is up regulated in many different tumour types and cell lines,14–16 often in areas of tumour hypoxia.17 Fibroblast growth factor 2 is ubiquitous in the extracellular matrix bound in an inactive form, but in tumours the active factor is released either from the tumour or from the extracellular matrix by various mechanisms.18 Increased serum and urine concentrations of fibroblast growth factor 2 can be detected in cancer patients.19 Increasing the amounts of active fibroblast growth factors secreted by tumour cell lines, by transfecting DNA into cells, can markedly increase the ability of poorly tumorigenic cells to form tumours and to metastasise.20
Because of the evidence that vascular endothelial growth factor and fibroblast growth factor 2 are promoters of tumour angiogenesis, researchers have investigated ways of interfering with their angiogenic effects and demonstrated the ability to induce tumour regression in mice. These approaches include targeting vascular endothelial growth factor with a monoclonal antibody21 and preventing release of active fibroblast growth factor 2 from the extracellular matrix by eradicating a binding protein necessary for its release.22 The problem with such approaches is that they inhibit only a single positive factor. There is a range of positive angiogenic factors, and many have been shown to be up regulated in tumours. More importantly, individual tumours can express several angiogenic factors23 and thus may have a number of routes around such a specific approach.
Angiogenic inhibitors as therapeutic agents
An alternative approach is to use an angiogenesis inhibitor in order to counter the sum effects of all angiogenic factors produced by a tumour. These agents can be synthetic inhibitors of endothelial cell proliferation—such as synthetic derivatives of fumagillin—or endogenous inhibitors of angiogenesis—often fragments of larger inactive circulatory proteins—that function physiologically in maintaining vascular quiescence or curtailing physiological angiogenesis. These endogenous agents can be highly potent inhibitors of endothelial proliferation and may be more effective than synthetic agents because they may also inhibit the capillary remodelling that is involved in expansion of tumour vessels. Angiostatin and endostatin are currently the most potent agents, both having striking antitumour activity. Others include a 140 kDa fragment of thrombospondin, a protein normally involved in platelet aggregation under the regulation of the tumour suppresser gene p53. 24 25
Angiostatin and endostatin
The discovery of these two potent endogenous inhibitors of angiogenesis with powerful antitumour activity in mice has produced great interest in the clinical use of angiogenesis inhibitors.26–28 Both were discovered as a result of the observation that the presence of a primary tumour can occasionally inhibit the development of metastases: when the primary tumour is removed, metastases develop rapidly. Folkman hypothesised that the primary tumour produced angiogenesis inhibitors, perhaps incidentally as a result of proteolytic degradation. These inhibitors persisted in the circulation while local angiogenic promoters were degraded and exerted no systemic effect. Angiostatin and endostatin were identified from experimental tumours that demonstrated this phenomenon.
Both are fragments of larger circulatory proteins with no angiogenic activity. Angiostatin is a 38 kDa fragment of plasminogen, and endostatin is a fragment of collagen XVIII, a type of collagen found exclusively in blood vessels. Their antitumour activity in mice is impressive, being able to cause regression in various solid tumours up to 1% of body mass. There is no evidence of drug resistance, even after multiple treatment cycles.29 Research efforts are currently directed at elucidating the precise mechanism of action of these agents, and at their large scale purification, before they are tested in clinical trials.
Targeting endothelial cells of tumours
Another experimental antivascular approach to treating tumours is to target tumour endothelial cells with a toxin directed at a cell marker specific for tumour endothelium and cause infarction of the tumour by inducing coagulation within the tumour vessels. This has been shown to be effective experimentally by using a neuroblastoma tumour that was genetically engineered to express the class II histocompatibility antigens on tumour endothelial cells. These antigens are normally absent from endothelial cells, so they served as specific markers for tumour vessels. A toxin was constructed consisting of an antibody to class II antigens linked to a truncated form of tissue factor that would cause coagulation only when bound to an endothelial cell by the antibody. This toxin induced complete regression of the experimental tumour by thrombosis of the tumour vessels while leaving the host vasculature intact.30
Potential endogenous target molecules include integrin αvβ3, which is expressed only on proliferating vessels in healing wounds and in tumours. Antibodies to integrin αvβ3 promote tumour regression by inducing endothelial cell apoptosis.31
Future developments
Preclinical research into tumour angiogenesis has led to the identification of several antivascular treatments with impressive efficacy in animal models of human cancer. Currently, 19 antivascular agents are being assessed in clinical trials, mostly still phase I and II trials that involve treating patients with advanced metastatic disease that is resistant to other treatments. (Information about current trials can be obtained from the National Cancer Institute's website at cancertrials.nci.nih.gov.) There are occasional reports of striking clinical remissions,32 but the real efficacy of these agents will only become apparent over the next decade as they are fully evaluated in extensive clinical studies either alone or with standard treatments.
As inhibition of angiogenesis may induce dormancy of a tumour rather than killing it, there is growing appreciation that the administration of these agents, and their assessment in clinical trials, may need to be different from that currently used for cytotoxic drugs. Indeed, it is possible that these agents could be effective in maintaining long term remission, an approach not currently used for solid tumours.
Footnotes
-
Competing interest None declared.