A perspective on vascular disrupting agents that interact with tubulin: preclinical tumor imaging and biological assessment

Abstract

The tumor microenvironment provides a rich source of potential targets for selective therapeutic intervention with properly designed anticancer agents. Significant physiological differences exist between the microvessels that nourish tumors and those that supply healthy tissue. Selective drug-mediated damage of these tortuous and chaotic microvessels starves a tumor of necessary nutrients and oxygen and eventually leads to massive tumor necrosis. Vascular targeting strategies in oncology are divided into two separate groups: angiogenesis inhibiting agents (AIAs) and vascular disrupting agents (VDAs). The mechanisms of action between these two classes of compounds are profoundly distinct. The AIAs inhibit the actual formation of new vessels, while the VDAs damage and/or destroy existing tumor vasculature. One subset of small-molecule VDAs functions by inhibiting the assembly of tubulin into microtubules, thus causing morphology changes to the endothelial cells lining the tumor vasculature, triggered by a cascade of cell signaling events. Ultimately this results in catastrophic damage to the vessels feeding the tumor. The rapid emergence and subsequent development of the VDA field over the past decade has led to the establishment of a synergistic combination of preclinical state-of-the-art tumor imaging and biological evaluation strategies that are often indicative of future clinical efficacy for a given VDA. This review focuses on an integration of the appropriate biochemical and biological tools necessary to assess (preclinically) new small-molecule, tubulin active VDAs for their potential to be clinically effective anticancer agents.

Fig. 1. Tubulin Binding Small-Molecule Vascular Disrupting Agents (VDAs)

A compilation of VDAs currently in human clinical trials that function, in part, through a tubulin mechanism include: CA4P-, AVE8062, CA1P-, MPC-6827, ABT-751, TZT-1027, CYT997,, MN-029, NPI-2358, BNC105P, EPC2407-, CKD-516.

Fig. 2

Model for downstream effects of treatment with tubulin-binding VDAs in tumor endothelium. Microtubule disassembly leads to cell rounding and detachment that can result in a rapid collapse in tumor blood flow.

Fig. 3

(A) Inhibition of tubulin polymerization as a function of increasing concentration of VDA (CA1) from the top two traces, 0 μM (control), to the lowest curve at 5 μM CA1. (B) IC₅₀ determination for CA1 (1.9 μM).

Fig. 4. Vascular heterogeneity in tumors identified ex vivo

These panels provide evidence for vascular heterogeneity based on ex vivo analysis:

1. A)

   Scanning Electron Microscopy reveals disorganized tumor vasculature with many tortuous microvessels and blind ends. A 13762NF mammary adenocarcinoma was grown using a kidney tissue-isolated preparation in a syngeneic adult female Fischer 344 rat. Prior to sacrifice the abdominal aorta proximal to the renal artery was perfused with Batson’s #17 polymer generating a vascular cast. The vascular corrosion cast was sputtered with gold-palladium prior to SEM (data obtained in collaboration with Andrew Abbott and Dr. C. Gilpin).

2. B)

   Immunohistochemistry reveals vascular extent in a Dunning prostate R3327-HI tumor based on anti-CD31 mAb. The tumor was grown in a syngeneic male Copenhagen rat and Hoechst 33342 dye (seen in blue in C) was infused IV followed by tumor excision 60 s later. The green stain shows vascular extent based on anti-CD31 mAb in a frozen 6 μm-thick section. Original magnification x100.

3. C)

   Extravasated Hoechst dye (blue) reveals tumor perfusion in same slice as B.

4. D)

   Overlay of the fluorescent images in B and C reveals that essentially all blood vessels were perfused in this tumor.

Fig. 5. Evaluating VDAs non-invasively by MRI

Dynamic contrast enhanced (DCE) MRI was performed pre, and 2 h and 24 h after treatment of a 13762NF rat breast tumor with CA4P (30 mg/kg, IP). A series of T₁-weighted images was acquired with respect to infusion of a bolus injection of contrast agent (Gd-DTPA-BMA (Omniscan); 0.1 mmol/kg). A) Normalized images of signal enhancement 25 s after contrast agent administration are shown for a representative tumor. Significantly less signal enhancement was observed for the whole tumor region 2 h after treatment with CA4P. A recovery was apparent in the tumor rim 24 h later. B) Time course curves showed rapid increase in signal intensity after infusion of contrast agent prior to CA4P, whereas much slower increase and significantly less enhancement was observed 2 h after CA4P. Twenty-four hours later signal enhancement recovered to ~ 90% of the baseline.

Fig. 6. Assessment of hypoxia accompanying vascular disruption

A) pO₂ maps obtained using FREDOM (¹⁹F MRI) with respect to hyperoxic gas breathing challenge and CA4P treatment (30 mg/kg, IP). Following administration of hexafluorobenzene (HFB) directly into the 13762NF tumor ten individual locations (voxels) could be examined sequentially with respect to an oxygen breathing and CA4P administration over a period up to 2 h. The next day, residual HFB allowed further interrogation of the tumor using FREDOM showing thirteen voxels, which could be assessed with respect to oxygen challenge. At baseline, there was a significant increase in pO₂ with respect to breathing oxygen. A significant decrease in pO₂ was evident for all the individual voxels after CA4P, and pO₂ no longer responded to oxygen inhalation after 2 h. The 24 h maps showed that pO₂ again responded to oxygen breathing. B) Mean tumor pO₂ with respect to interventions. * p < 0.05 from baseline air, ⁺ p <0.05 from 24 h air.

Fig. 7. Assessment of acute changes in tumor vasculature using ultrasound

High resolution ultrasound using a VisualSonics Vevo770 small animal ultrasound device of an 8 mm diameter syngeneic MTLn3 mammary tumor growing subcutaneously in a Fisher rat with respect to infusion of the new VDA CA1P (30 mg/kg, IP) A) Doppler mode using 40 MHz transducer shows a few major blood vessels in tumor and significant vasculature in skin. B) Infusion of commercial contrast microbubbles revealed vasculature more clearly, particularly revealing small blood vessels using a 40 MHz transducer. C) 30 mins after infusion of CA1P the Color Doppler indicated minor reduction in blood flow. D) Infusion of fresh contrast agent clearly showed substantial decrease in perfusion.

Fig. 8. Assessment of acute changes in tumor vasculature using dynamic bioluminescent imaging (dBLI)

1. A)

   Human prostate PC3-Luc cells were implanted subcutaneously in the right thighs of two nude mice and observed by BLI following administration of luciferin 80 μl (40 mg/ml) SC in the shoulder neck region. Images were acquired every 60 s for 30 mins to observe the dynamic evolution of the bioluminescent signal (dynamic BLI). Following a baseline time course, CA4P was administered IP (100 μl; 150 mg/kg) and 2 h later dBLI was repeated with fresh luciferin. DBLI was repeated the following day.

2. B)

   Mean light intensity indicated a maximum at 25 min under control conditions. Following CA4P light emission was decreased at least 90%. Twenty-four hours later there was substantial recovery in light emission reaching about 30% of baseline.