Alastair H. Kyle

Angiopoietin-2 Functions as a Tie2 Agonist in Tumor Models, Where It Limits the Effects of VEGF Inhibition

The angiopoietins Ang1 (ANGPT1) and Ang2 (ANGPT2) are secreted factors that bind to the endothelial cell-specific receptor tyrosine kinase Tie2 (TEK) and regulate angiogenesis. Ang1 activates Tie2 to promote blood vessel maturation and stabilization. In contrast, Ang2, which is highly expressed by tumor endothelial cells, is thought to inhibit Tie2 activity and destabilize blood vessels, thereby facilitating VEGF-dependent vessel growth. Here, we show that the inhibition of tumor xenograft growth caused by an Ang2-specific antibody (REGN910) is reversed by systemic administration of the Tie2 agonist Ang1. These results indicate that Ang2 blockade inhibits tumor growth by decreasing Tie2 activity, showing that Ang2 is a Tie2 activator. REGN910 treatment of tumors resulted in increased expression of genes that are repressed by Tie2 activation, providing further evidence that REGN910 inhibits Tie2 signaling. Combination treatment with REGN910 plus the VEGF blocker aflibercept reduced tumor vascularity and tumor perfusion more dramatically than either single agent, resulting in more extensive tumor cell death and more potent inhibition of tumor growth. Challenging the prevailing model of Ang2 as a destabilizing factor, our findings indicate that Ang2 plays a protective role in tumor endothelial cells by activating Tie2, thereby limiting the antivascular effects of VEGF inhibition. Thus, blockade of Ang2 might enhance the clinical benefits currently provided by anti-VEGF agents. .

Tissue Penetration and Activity of Camptothecins in Solid Tumor Xenografts

The ability of a panel of camptothecin derivatives to access the tumor compartment was evaluated to determine the mechanisms by which the architecture of solid tumors may act to limit their activity. Microregional localization and activity of members of the camptothecin class of topoisomerase I targeting agents, including topotecan, irinotecan, and irinophore C, a lipid-based nanoparticulate formulation of irinotecan, were evaluated over time in HCT116 and HT29 colorectal tumor xenografts. Using native drug fluorescence, their distributions in tissue cryosections were related to the underlying tumor vasculature, tumor cell proliferation, and apoptosis. Topotecan exhibited a relatively uniform tumor distribution; in tissue 100 μm away from vessels, it reached 94% ± 5% of levels seen around blood vessels, whereas irinotecan and irinophore C were found to reach only 41% ± 10% and 5% ± 2%, respectively. Surprisingly, all three agents were able to initially inhibit proliferation uniformly throughout the tumors, and it was their rate of washout (topotecan > irinotecan > irinophore C) that correlated with activity. To explain this discrepancy, we looked at SN38, the active metabolite of irinotecan, and found it to penetrate tissue similarly to topotecan. Hence, the poor access to the tumor compartment of irinotecan and irinophore C could be offset by their systemic conversion to SN38. It was concluded that all three agents were effective at reaching tumor cells, and that despite the poor access to the extravascular compartment of irinophore C, its extended plasma exposure and systemic conversion to the diffusible metabolite SN38 enabled it to effectively target solid tumors. Mol Cancer Ther; 13(11); 2727–37. ©2014 AACR.

Targeting the tumour vasculature: exploitation of low oxygenation and sensitivity to NOS inhibition by treatment with a hypoxic cytotoxin.

Many cancer research efforts focus on exploiting genetic-level features that may be targeted for therapy. Tissue-level features of the tumour microenvironment also represent useful therapeutic targets. Here we investigate the presence of low oxygen tension and sensitivity to NOS inhibition of tumour vasculature as potential tumour-specific features that may be targeted by hypoxic cytotoxins, a class of therapeutics currently under investigation. We have previously demonstrated that tirapazamine (TPZ) mediates central vascular dysfunction in tumours. TPZ is a hypoxic cytotoxin that is also a competitive inhibitor of NOS. Here we further investigated the vascular-targeting activity of TPZ by combining it with NOS inhibitor L-NNA, or with low oxygen content gas breathing. Tumours were analyzed via multiplex immunohistochemical staining that revealed irreversible loss of perfusion and enhanced tumour cell death when TPZ was combined with either low oxygen or a NOS inhibitor. Tumour growth rate was reduced by TPZ + NOS inhibition, and tumours previously resistant to TPZ-mediated vascular dysfunction were sensitized by low oxygen breathing. Additional mapping analysis suggests that tumours with reduced vascular-associated stroma may have greater sensitivity to these effects. These results indicate that poorly oxygenated tumour vessels, also being abnormally organized and with inadequate smooth muscle, may be successfully targeted for significant anti-cancer effects by inhibition of NOS and hypoxia-activated prodrug toxicity. This strategy illustrates a novel use of hypoxia-activated cytotoxic prodrugs as vascular targeting agents, and also represents a novel mechanism for targeting tumour vessels.

Heterogeneous distribution of trastuzumab in HER2-positive xenografts and metastases: role of the tumor microenvironment

Most HER2-positive metastatic breast cancer patients continue to relapse. Incomplete access to all target HER2-positive cells in metastases and tumor tissues is a potential mechanism of resistance to trastuzumab. The location of locally bound trastuzumab was evaluated in HER2-positive tissues in vivo and as in vivo xenografts or metastases models in mice. Microenvironmental elements of tumors were related to bound trastuzumab using immunohistochemical staining and include tight junctions, vasculature, vascular maturity, vessel patency, hypoxia and HER2 to look for correlations. Trastuzumab was evaluated alone and in combination with bevacizumab. Dynamic contrast-enhanced magnetic resonance imaging parameters of overall vascular function, perfusion and apparent permeability were compared with matched histological images of trastuzumab distribution and vascular patency. Trastuzumab distribution is highly heterogeneous in all models examined, including avascular micrometastases of the brain and lung. Trastuzumab distributes well through the extravascular compartment even in conditions of high HER2 expression and poor convective flow in vivo. Microregional patterns of trastuzumab distribution in vivo do not consistently correlate with vascular density, patency, function or maturity; areas of poor trastuzumab access are not necessarily those with poor vascular supply. The number of vessels with perivascular trastuzumab increases with time and higher doses and dramatically decreases when pre-treated with bevacizumab. Areas of HER2-positive tissue without bound trastuzumab persist in all conditions. These data directly demonstrate tissue- and vessel-level barriers to trastuzumab distribution in vivo that can effectively limit access of the drug to target cells in brain metastases and elsewhere.

Abstract B32: Microenvironmental distribution of trastuzumab in metastases and xenograft models is highly heterogeneous and decreases sharply when administered in combination with bevacizumab

Despite significant success, the response of Her2+ patients to trastuzumab (TzMAb, Herceptin ®) is varied, with many still experiencing tumor progression. We have used 3D tissue, tumor xenograft and metastatic models to examine the microregional distribution of TzMAb when administered alone or in combination with bevacizumab (BvMAb). Methods: Her2 positive (Her2+) SKOV-3 ovarian and MDA361, JIMT-1, BT474 mammary cancer cells were grown as 3D tissue discs, spheroids and as xenografts. Her2 expression was ranked as SKOV3 > BT474 > JIMT-1 > MDA361. Variable concentrations (25-100 µg/mL for in vitro; 2.5-10 mg/kg ip q3d for in vivo) of TzMAb, BvMAb or isotype IgG control antibodies were administered and tissues collected. Multiplexed immunohistochemistry generated maps of whole tissue sections for quantitative and qualitative analysis. In addition to direct visualization of fluorescent-tagged antibody therapeutics, features including Her2 expression (Her2), hypoxia (pimonidazole), blood vessels (CD31), vascular perfusion (carbocyanine fluorescent dye), pericytes (SMA, desmin), basal lamina (CIV) and tight junctions (ZO-1) were mapped relative to each other. Findings: In vitro: The rate of TzMAb distribution through 3D Her2+ tissue models was not significantly different to that of BvMAb or an isotype control (IgG). Inter-model variability was not correlated with the degree of Her2 expression. In vivo: All the xenograft models exhibited highly heterogeneous distribution of TzMAb, with variation at inter-vessel, inter-tumor and intra-tumor levels. There was no discernible pattern in the deposition of TzMAb; it was not limited to tumor margins or the central core, and was often bound up to 200 µm away from the nearest blood vessels. Other areas containing perfused vessels had little to no binding. Areas of limited TzMAb binding persisted after repeat dosing. SKOV-3 and BT474 metastatic lesions collected from liver and lung tissues had on average greater TzMAb staining intensity than was seen in subcutaneous tumors. However, there was a significant range in the ability of TzMAb to access and bind Her2+ve cells. Lesions ranged from less than 150 µm to several mm in diameter, with some metastases showing bound TzMAb on every cell and others showing no observable bound drug despite containing perfused vessels or being surrounded by highly perfused normal lung or liver tissue. For both subcutaneous tumors and metastatic lesions, no consistent, quantifiable difference or correlation was found between the amount of bound TzMAb and the presence of tight junctions, the density of nearest CD31 vessels, the location of hypoxic cells or the fraction of perfused, mature or immature vessels. Pretreatment with BvMAb for 24-72h resulted in significant reductions (up to 90% reduced) in tumor accumulation of TzMAb in all tumor models examined, including lung and liver metastases, despite the continued presence of perfused vessels. Summary: The extravascular distribution of TzMAb in Her2-overexpressing xenografts and metastases is heterogeneous and is not explained simply by the presence or absence of functioning vasculature. The relatively good distribution of TzMAb through 3D tissue models suggest that tight or irreversible binding to Her2 may not adequately explain the limited distribution of TzMAb seen in vivo. Pre-treatment with BvMAb dramatically decreases the access of TzMAb to the tumor microenvironment. Persistence of metastases and tissues with poor access to TzMAb suggests the possibility of inadequate drug exposure as a mechanism for resistance to TzMAb activity that could apply to additional targeted monoclonal antibody therapeutics. This resistance may be exacerbated when TzMAb or other antibodies are administered in combination with anti VEGF therapy. Citation Format: Jennifer H.E. Baker, Alastair H. Kyle, Jordan Cran, Haley Patrick, Maria-Jose Gandolfo, Andrew I. Minchinton. Microenvironmental distribution of trastuzumab in metastases and xenograft models is highly heterogeneous and decreases sharply when administered in combination with bevacizumab. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Breast Cancer Research; Oct 17-20, 2015; Bellevue, WA. Philadelphia (PA): AACR; Mol Cancer Res 2016;14(2_Suppl):Abstract nr B32.

Tissue Discs: A 3D Model for Assessing Modulation of Tissue Oxygenation

The presence of hypoxia in solid tumours is correlated with poor treatment outcome. We have developed a 3-D tissue engineered construct to quantitatively monitor oxygen penetration through tumour tissue using the exogenous 2-nitroimidazole bioreductive probe pimonidazole and phosphorescence quenching technologies. Using this in vitro model we were able to examine the influence of the biguanides metformin and phenformin, antimycin A and KCN, on the distribution and kinetics of oxygen delivery as prototypes of modulators of oxygen metabolism.

Abstract A12: Heterogeneous accumulation of trastuzumab in Her2+ve tumor and metastases models

Despite significant success, the response of Her2+ patients to trastuzumab (Herceptin ®) is varied, with many still experiencing tumour progression. In previous studies we have shown that trastuzumab has limited distribution in MDA-435-LCC6-Her2 over-expressing breast cancer xenografts. We have used biomarkers derived from MRI and multiplexed immunohistochemistry to examine the limited distribution of trastuzumab in the context of the tumour microenvironment. Methods: Endogenously over-expressing SKOV-3 ovarian carcinomas, MDA361, BT474 and vector-overexpressed Her2+ve MCF7 mammary (MCF7Her2) tumours were grown as xenografts in NOD/SCID mice in the sacral region, or were injected ip or iv to generate metastases in the peritoneal or lung compartments. Animals were administered 10 mg/kg trastuzumab with tissues harvested & frozen at 20-24h. MRI: Tumours were imaged for vascular function using DCE-MRI (Gadovist, 60mM), then treated with trastuzumab. Implanted fiducial markers enabled cryosections that closely approximated MR imaged slices to be obtained for comparison to tumour maps. Area Under the Curve (AUC) was calculated for the first 60 seconds post contrast agent injection. Tumour mapping: Multiplexed immunohistochemistry generated maps of whole tumour sections for quantitative and qualitative analysis. In addition to direct visualization of trastuzumab, features including Her2 expression (Her2), blood vessels (CD31), vascular perfusion (Dil18) and pericytes (SMA, desmin), basal lamina (CIV) and tight junctions (ZO-1) were stained and mapped relative to each other. Findings: All models exhibit highly heterogeneous distribution of trastuzumab, with variation at the inter-vessel, inter-tumour and intra-tumoural levels. Accumulation of trastuzumab is not limited to the tumour margins, though sections obtained on the distal ends of tumours have greater average amounts of bound drug. Trastuzumab is bound to tumour cells up to 200 µm away from the nearest perfused blood vessels in some areas, while other areas containing perfused vessels have little to no binding despite Her2 expression. The average amount of Her2+ve tissue positive for bound trastuzumab at 20-24h post administration is: MCF7-Her2 (68.6±7.2%), SKOV-3 (31.5±1.9%), BT474 (33.2%±1.6), MDA361 (21.5%+3.2). Small SKOV-3 metastatic lesions collected from the peritoneal cavity and lung tissues have an average of 68.5±3.5% of Her2+ve tissue bound for trastuzumab, although many have as little as 30%, and some very small lesions have no bound trastuzumab at all. For both the solid tumours and the metastatic lesions, no consistent, quantifiable difference or correlation was found between the amount of bound trastuzumab and the presence of greater ZO-1 labeled tight junctions, the density of nearest CD31 vessels, or the fraction of perfused, desmin +ve or SMA +ve vessels. Maps derived from DCE-MRI data showed that in some cases a high AUC value matched with high trastuzumab saturation, however these patterns were not consistent; in every tumour significant areas of mismatched high and low regions were observed. Summary: The highly heterogeneous microregional distribution of trastuzumab in these models appears to be independent of conventionally considered barriers to drug delivery, including ZO-1 tight junctions, microvessel density, blood flow and vascular maturity. The common occurrence of perfused vessels with no trastuzumab on the perivacular Her2+ve cells suggests that limited distribution is not a consequence of the binding site barrier hypothesis. The absence of a consistent pattern of greater accumulation of trastuzumab at the tumour margins suggests that interstitial fluid pressure (IFP) is not primarily responsible for the observed heterogeneity and limited distribution. These data suggest an as yet unidentified barrier may be responsible for inadequate access of trastuzumab to all Her2+ve cells. Citation Format: Jennifer Baker, Alastair H. Kyle, Firas Moosvi, Andrew I. Minchinton. Heterogeneous accumulation of trastuzumab in Her2+ve tumor and metastases models. [abstract]. In: Abstracts: AACR Special Conference on Cellular Heterogeneity in the Tumor Microenvironment; 2014 Feb 26-Mar 1; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(1 Suppl):Abstract nr A12. doi:10.1158/1538-7445.CHTME14-A12

Abstract 4314: Metabolic manipulation of hypoxia and radiotherapy response by electron transport inhibitors

Proceedings: AACR Annual Meeting 2014; April 5-9, 2014; San Diego, CAnnThe presence of hypoxia in solid tumors is correlated with poor treatment outcome. Even small population of hypoxic cells within tumors can significantly decrease the efficacy of radiation therapy. One approach to treating hypoxia in tumours is to reduce overall tumor cell oxygen consumption by modulating mitochondrial respiration. This could reoxygenate hypoxic cells within the tumor and restore radiation sensitivity. To guide in vivo application we tested the effects of three respiratory chain inhibitors; the biguanides metformin and phenformin and rotenone on oxygen consumption in 3D tissue engineered discs.nnExperiments were carried out to evaluate the effect of respiratory inhibition on the hypoxic fraction of HCT116 and HT29 tissue discs using immunostaining with the hypoxic cell marker pimonidazole. An oxygen sensing phosphorescence lifetime probe assessed pO2 on the basal side of the disc as a function of time after inhibitors were exposed to the apical surface. Clonogenic assays were then performed to confirm the radiosensitization of hypoxic cells within the tissue discs.nnPimonidazole staining and direct pO2 sensing confirmed that all three electron transfer inhibitors decrease the hypoxic fraction by reducing cellular oxygen consumption. Rotenone was found to completely eliminate the hypoxic fraction at nanomolar concentrations compared to metformin and phenformin which require milli to micromolar concentrations respectively. Furthermore, rotenone was found to have an effect on oxygen utilization within minutes of drug exposure compared to both metformin and phenformin, which took hours to be effective. In addition, clonogenics assays determined that all three electron transfer inhibitors radiosensitize tumor cells with rotenone being the most efficient compound on a molar basis.nnThe mitochondrial inhibitors of complex I metformin, phenformin and rotenone decrease oxygen utilization by tumor cells in a concentration dependent manner resulting in reduction of the hypoxic fraction and increased radiosensitivity.nnCitation Format: Maria Jose Gandolfo, Alastair H. Kyle, Andrew I. Minchinton. Metabolic manipulation of hypoxia and radiotherapy response by electron transport inhibitors. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4314. doi:10.1158/1538-7445.AM2014-4314

Abstract PR-7: Overcoming chemoresistance in solid cancers by targeting quiescent tumor cells

Overview: Quiescent tumor cells in solid cancers contribute significantly to treatment failure of conventional anti‐proliferative based therapies. The hypothesis of this study is that transiently increasing the proliferative fraction of a tumor will increase response to chemotherapy and result in a net benefit to overall survival. Initial work was carried out using an in vitro engineered‐tissue assay to develop and test strategies to induce proliferation. Using this assay a combination of growth factor stimulation and O2 supplementation was found to override tumor‐derived quiescence and enhance response to a panel of chemotherapy drugs. Background: The microenvironment in solid tumors limits the efficacy of anti‐proliferative therapies by generating a quiescent tumor cell subpopulation through diffusion‐limited supply of nutrients supplied by the blood. Recent findings in our laboratory have shown that the distribution of proliferating cells in solid tumors can be modified if the limiting factors controlling it are understood. Quiescence in tissues grown from two human colorectal cancer cell lines was shown to be driven by oxygen and growth factor deprivation. The combined supplementation of these two factors was found to reverse tumor microenvironment derived quiescence. Methods: An in vitro engineered‐tissue model grown from HCT116 & HT29 cells was used to study the interrelation of key factors involved in limiting tumor cell proliferation. The model was able to replicate gradients in diffusible factors and reproduce quiescence seen in solid tumors. Findings were validated in tumor xenografts using a technique in which proliferation was mapped in relation to tumor vasculature. Strategies to stimulate proliferation in combination with treatment were then tested against a panel of drugs with cell survival as the endpoint (doxorubicin, paclitaxel, vinorelbine, gemcitabine & SN38). Results: Stimulation of the IGF‐1 receptor in combination with O2 supplementation in the engineered‐tissue model was able to initiate proliferation in quiescent areas with maximal induction occurring after 16–24 hours, tissue proliferation increased 3±0.4‐fold (HCT116) and 4±1‐fold (HT29) as measured using a BrdUrd endpoint. Chemotherapy given at the time of maximal induction induced a 3 to 6‐fold increase in cell kill when compared to tissue grown under normal physiological conditions. Findings showed that combined stimulation always yielded a greater effect than individual stimulation via IGF‐1R or oxygen alone. Cell kill was also greater when chemotherapy was given 16 hours following initiation of stimulation compared to when given simultaneously. Typical results: 1µM gemcitabine with no induction SF=0.71±0.08, with IGF‐1R stimulation SF=0.25±0.03, with O2 supplementation SF=0.47±0.04, with combined induction 12 hours prior SF=0.10±0.03, with combined induction at time SF=0.43±0.03 (mean ± SD, n=8). Conclusions: Findings from this study indicate that transient stimulation of quiescent tumor cells represents a promising target for improving the activity of most standard chemotherapy regimes. The in vitro engineered‐tissue model was found to be a powerful tool to study diffusion limited supply of molecules in solid tumors. Initial results using the engineered‐tissue model have been validated in tumor xenografts and further work to translate these findings is underway. Citation Information: Mol Cancer Ther 2009;8(12 Suppl):PR-7.