Sally A. Hill

The biology of the combretastatins as tumour vascular targeting agents

The tumour vasculature is an attractive target for therapy. Combretastatin A‐4 (CA‐4) and A‐1 (CA‐1) are tubulin binding agents, structurally related to colchicine, which induce vascular‐mediated tumour necrosis in animal models. CA‐1 and CA‐4 were isolated from the African bush willow, Combretum caffrum, and several synthetic analogues are also now available, such as the Aventis Pharma compound, AVE8062. More soluble, phosphated, forms of CA‐4 (CA‐4‐P) and CA‐1 (CA‐1‐P) are commonly used for in vitro and in vivo studies. These are cleaved to the natural forms by endogenous phosphatases and are taken up into cells. The lead compound, CA‐4‐P, is currently in clinical trial as a tumour vascular targeting agent. In animal models, CA‐4‐P causes a prolonged and extensive shut‐down of blood flow in established tumour blood vessels, with much less effect in normal tissues. This paper reviews the current understanding of the mechanism of action of the combretastatins and their therapeutic potential.

Vinca alkaloids: Anti-vascular effects in a murine tumour

We have investigated the blood flow modifying effects of the vinca alkaloids, vincristine and vinblastine in the murine carcinoma CaNT. Vinblastine at doses of 7.5 or 10 mg/kg induced profound and chronic reductions in tumour blood flow as measured by 86RbCl extraction. Following the maximum tolerated dose of 10 mg/kg, blood flow was reduced to 10% of pretreatment values after 2 h and remained below 20% of pretreatment values 24 h after drug administration. These findings are consistent with the early induction of necrosis by vinblastine and suggest that vascular-mediated cell death may account for a large part of the 11 day growth delay induced by this drug dose. In contrast to the large reductions in tumour blood flow, in skin, kidney, liver and muscle, blood flow reductions did not, at any time examined, exceed 40%. In all the normal tissues studied, blood flow had fully recovered by 6 h after vinblastine administration. Similar results, albeit less pronounced, have been obtained with vincristine at the maximum tolerated dose of 3 mg/kg. The results clearly show that both vinblastine and vincristine can induce, with some selectivity, a dramatic and prolonged reduction in tumour blood flow and that this may contribute to the anti-tumour effects against the CaNT tumour.

Vascular occlusion and tumour cell death

Vascular occlusion has been tested as a means of inducing regrowth delay, local control, reduced cell viability and prolonged alteration of blood flow in mouse tumours. The occlusion has been achieved by applying D-shaped metal clamps across the base of subcutaneously implanted tumours. The period of clamping has been varied from 30 min to 24 hr. Marked tumour regression, delayed growth and long-term tumour control were seen, with the magnitude of the response being proportional to the duration of clamping. Vessel occlusion for at least 15 hr is necessary to achieve local cure of the tumour. The overall effect results partly from an immediate loss of cell viability and partly from a failure of the capillary network to recover its normal perfusion pattern after the clamp has been removed. The implications of this for anti-proliferative endothelial therapy is discussed.

THE DEVELOPMENT OF COMBRETASTATIN A4 PHOSPHATE AS A VASCULAR TARGETING AGENT

PURPOSE This overview summarizes the preclinical development of tubulin-depolymerizing agents as vascular targeting agents, leading to the identification of combretastatin A4P (CA4P). METHODS AND MATERIALS The murine tumor CaNT was implanted s.c. in the dorsum of CBA mice. Vascular function was determined after treatment using the perfusion marker Hoechst 33342 and fluorescence microscopy. Tumor cell response was assessed by using an excision assay and by measuring the delay in growth of treated tumors. RESULTS At doses that approximated one-half the maximum tolerated dose (MTD) in CBA mice, none of the agents evaluated-i.e., taxol, melphalan, 5-fluorouracil, doxorubicin, cisplatin, gemcitabine, and irinotecan-induced any significant reduction in perfused vascular volume within the tumor mass. In contrast, CA4P at a dose of 100 mg/kg, which approximates one-fifth the MTD, induced a greater than 80% reduction in vascular function. Although colchicine did induce vascular shutdown, this occurred only at doses approximating the MTD. Histologic evaluation demonstrated that continued growth and repopulation of the tumor mass was the result of a surviving rim of viable tumor cells at the tumor periphery. CONCLUSION These results confirm the ability of CA4P to selectively compromise vascular function in experimental tumors, inducing extensive tumor cell death at well-tolerated doses. However, despite these effects, no growth retardation is obtained when CA4P is administered alone in a single dose. The continued growth and repopulation of the tumor mass occurs from a narrow rim of viable cells at the periphery. If, as is believed, these remaining cells are the ones most sensitive to conventional cytotoxic and macromolecular approaches, CA4P and other vascular targeting agents offer considerable potential for enhancing the effectiveness of existing and emerging cancer therapies.

Magnetic resonance imaging and spectroscopy of combretastatin A4 prodrug-induced disruption of tumour perfusion and energetic status

The effects of combretastatin A4 prodrug on perfusion and the levels of 31P metabolites in an implanted murine tumour were investigated for 3 h after drug treatment using nuclear magnetic resonance imaging (MRI) and spectroscopy (MRS). The area of regions of low signal intensity in spin-echo images of tumours increased slightly after treatment with the drug. These regions of low signal intensity corresponded to necrosis seen in histological sections, whereas the expanding regions surrounding them corresponded to haemorrhage. Tumour perfusion was assessed before and 160 min after drug treatment using dynamic MRI measurements of gadolinium diethylenetriaminepentaacetate (GdDTPA) uptake and washout. Perfusion decreased significantly in central regions of the tumour after treatment. This was attributed to disruption of the vasculature and was consistent with the haemorrhage seen in histological sections. The mean apparent diffusion coefficient of water within the tumour did not change, indicating that there was no expansion of necrotic regions during the 3 h after drug treatment. Localized 31P-MRS showed that there was decline in cellular energy status in the tumour after treatment with the drug. The concentrations of nucleoside triphosphates within the tumour fell, the inorganic phosphate concentration increased and there was a significant decrease in tumour pH for 80 min after drug treatment. The rapid, selective and extensive damage caused to these tumours by combretastatin A4 prodrug has highlighted the potential of the agent as a novel cancer chemotherapeutic agent. We have shown that the response of tumours to treatment with the drug may be monitored non-invasively using MRI and MRS experiments that are appropriate for use in a clinical setting.

Validation of the fluorescent dye Hoechst 33342 as a vascular space marker in tumours.

The DNA-binding fluorescent dye Hoechst 33342 (H33342) has been used in a series of investigations of the vascular parameters of two murine tumours. This dye has been shown, to have a short half-life in the circulation (T1/2 less than 2 min), but is stably bound for at least 2 h after it enters cells. It can be used in morphometric studies on frozen sections to determine the effective vascular volume, the capillary fraction and the size distribution of blood vessels in each tumour. These latter two parameters cannot be deduced from the less labour intensive techniques using radioactive isotopes. The effective vascular volume perfused in 1 min by H33342 was compared with the volume perfused in 30 min with 51Cr labelled erythrocytes. Similar volumes were estimated with the two techniques in a murine carcinoma and in a sarcoma. Both techniques showed that the vascular volume decreased in larger tumours. The H33342 analysis of vessel size showed the decrease in capillary vessels in the carcinomas was even greater, falling from 70% in small tumours to 20% in larger tumours. The deteriorating vascular network in larger tumours is associated with an increasing fraction of necrotic tissue. Experiments in which the isotopes and dye were co-injected suggest that at 40 mgkg-1 the dye may rapidly lead to a partial shutdown of the tumour vascular bed. This is less marked with 20 mg kg-1. In spite of this effect there is in general a close correlation between the volumes perfused by labelled red blood cells and the fluorescent dye.

Dual Inhibition of the PI3K/mTOR Pathway Increases Tumor Radiosensitivity by Normalizing Tumor Vasculature

The aberrant vascular architecture of solid tumors results in hypoxia that limits the efficacy of radiotherapy. Vascular normalization using antiangiogenic agents has been proposed as a means to improve radiation therapy by enhancing tumor oxygenation, but only short-lived effects for this strategy have been reported so far. Here, we show that NVP-BEZ235, a dual inhibitor of phosphoinositide-3-kinase (PI3K) and mTOR, can improve tumor oxygenation and vascular structure over a prolonged period that achieves the aim of effective vascular normalization. Because PI3K inhibition can radiosensitize tumor cells themselves, our experimental design explicitly distinguished effects on the blood vasculature versus tumor cells. Drug administration coincident with radiation enhanced the delay in tumor growth without changing tumor oxygenation, establishing that radiosensitization is a component of the response. However, the enhanced growth delay was substantially greater after induction of vascular normalization, meaning that this treatment enhanced the tumoral radioresponse. Importantly, changes in vascular morphology persisted throughout the entire course of the experiment. Our findings indicated that targeting the PI3K/mTOR pathway can modulate the tumor microenvironment to induce a prolonged normalization of blood vessels. The substantial therapeutic gain observed after combination of NVP-BEZ235 with irradiation has conceptual implications for cancer therapy and could be of broad translational importance.

Measurement of Tumor Oxygenation: A Comparison between Polarographic Needle Electrodes and a Time-Resolved Luminescence-Based Optical Sensor

A novel oxygen sensor which does not rely on electrochemical reduction has been used to measure the oxygenation of the murine sarcoma F in a comparative study with an existing polarographic electrode that is available commercially. The prototype luminescence sensor yielded an oxygen distribution comparable with readings made using a pO2 histograph. The percentage of regions detected that had a pO2 less than 5 mm Hg was 79 and 75 using the Eppendorf pO2 histograph and the luminescence fiber optic sensor, respectively. These values were compatible with a measured radiobiologically hypoxic fraction of 67% in this tumor. The polarographic method detected more regions with a pO2 of 2.5 mm Hg or less (69%) compared with the optical sensor (50%) (P < 0.05). This could reflect differences in the oxygen use of the sensing devices. This initial assessment indicates the potential of a fiber-optic-based oxygen-monitoring system. Such a system should have several advantages including monitoring temporal oxygen changes in a given microregion and use with NMR procedures.

Blood vessel maturation and response to vascular-disrupting therapy in single vascular endothelial growth factor-A isoform-producing tumors

Tubulin-binding vascular-disrupting agents (VDA) are currently in clinical trials for cancer therapy but the factors that influence tumor susceptibility to these agents are poorly understood. We evaluated the consequences of modifying tumor vascular morphology and function on vascular and therapeutic response to combretastatin-A4 3-O-phosphate (CA-4-P), which was chosen as a model VDA. Mouse fibrosarcoma cell lines that are capable of expressing all vascular endothelial growth factor (VEGF) isoforms (control) or only single isoforms of VEGF (VEGF120, VEGF164, or VEGF188) were developed under endogenous VEGF promoter control. Once tumors were established, VEGF isoform expression did not affect growth or blood flow rate. However, VEGF188 was uniquely associated with tumor vascular maturity, resistance to hemorrhage, and resistance to CA-4-P. Pericyte staining was much greater in VEGF188 and control tumors than in VEGF120 and VEGF164 tumors. Vascular volume was highest in VEGF120 and control tumors (CD31 staining) but total vascular length was highest in VEGF188 tumors, reflecting very narrow vessels forming complex vascular networks. I.v. administered 40 kDa FITC-dextran leaked slowly from the vasculature of VEGF188 tumors compared with VEGF120 tumors. Intravital microscopy measurements of vascular length and RBC velocity showed that CA-4-P produced significantly more vascular damage in VEGF120 and VEGF164 tumors than in VEGF188 and control tumors. Importantly, this translated into a similar differential in therapeutic response, as determined by tumor growth delay. Results imply differences in signaling pathways between VEGF isoforms and suggest that VEGF isoforms might be useful in vascular-disrupting cancer therapy to predict tumor susceptibility to VDAs.

Microregional fluctuations in perfusion within human tumours detected using laser Doppler flowmetry.

PURPOSE Transient fluctations in erythrocyte flux consistent with perfusion driven hypoxia have been previously reported using experimental tumour models. The present study was designed to establish whether such changes are a common feature of human tumours. METHODS AND MATERIALS A multi-channel laser Doppler system was used to monitor microregional changes in flow in human tumours. Eight individual tumours were investigated, two primary and one locally recurrent breast carcinoma, two metastatic skin deposits and three metastatic lymph nodes. Six custom designed microprobes (diameter of 300 microns), each monitoring a nominal sampling volume of approximately 10(-2) mm3 were inserted into the tumour and perfusion monitored over a period of 60 min. RESULTS The results show that in 54% of the regions monitored there was a change in microregional blood flow by a factor of 1.5 or more. Over the whole 60-min period, 19% of the changes were reversed, with a time course of 4-44 min. CONCLUSIONS This finding demonstrates that microregional fluctuations in perfusion occur frequently in human tumours. Furthermore, the observation that 19% of the changes were reversed implies that at least some of the cells are subject to transient acute hypoxia.