Open Collections

UBC Faculty Research and Publications

Vascular normalization in orthotopic glioblastoma following intravenous treatment with lipid-based nanoparticulate… Verreault, Maite; Strutt, Dita; Masin, Dana; Anantha, Malathi; Yung, Andrew; Kozlowski, Piotr; Waterhouse, Dawn; Bally, Marcel B; Yapp, Donald T Apr 8, 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12885_2010_Article_2616.pdf [ 4.22MB ]
JSON: 52383-1.0215895.json
JSON-LD: 52383-1.0215895-ld.json
RDF/XML (Pretty): 52383-1.0215895-rdf.xml
RDF/JSON: 52383-1.0215895-rdf.json
Turtle: 52383-1.0215895-turtle.txt
N-Triples: 52383-1.0215895-rdf-ntriples.txt
Original Record: 52383-1.0215895-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessVascular normalization in orthotopic glioblastomafollowing intravenous treatment with lipid-basednanoparticulate formulations of irinotecan(Irinophore C™), doxorubicin (Caelyx®) orvincristineMaite Verreault1*, Dita Strutt1, Dana Masin1, Malathi Anantha1, Andrew Yung5, Piotr Kozlowski5,Dawn Waterhouse1, Marcel B Bally1,2,3,4 and Donald T Yapp1,2AbstractBackground: Chemotherapy for glioblastoma (GBM) patients is compromised in part by poor perfusion in thetumor. The present study evaluates how treatment with liposomal formulation of irinotecan (Irinophore C™), andother liposomal anticancer drugs, influence the tumor vasculature of GBM models grown either orthotopically orsubcutaneously.Methods: Liposomal vincristine (2 mg/kg), doxorubicin (Caelyx®; 15 mg/kg) and irinotecan (Irinophore C™; 25 mg/kg) were injected intravenously (i.v.; once weekly for 3 weeks) in Rag2M mice bearing U251MG tumors. Tumorblood vessel function was assessed using the marker Hoechst 33342 and by magnetic resonance imaging-measured changes in vascular permeability/flow (Ktrans). Changes in CD31 staining density, basement membraneintegrity, pericyte coverage, blood vessel diameter were also assessed.Results: The three liposomal drugs inhibited tumor growth significantly compared to untreated control (p < 0.05-0.001). The effects on the tumor vasculature were determined 7 days following the last drug dose. There was a 2-3fold increase in the delivery of Hoechst 33342 observed in subcutaneous tumors (p < 0.001). In contrast there wasa 5-10 fold lower level of Hoechst 33342 delivery in the orthotopic model (p < 0.01), with the greatest effectobserved following treatment with Irinophore C. Following treatment with Irinophore C, there was a significantreduction in Ktrans in the orthotopic tumors (p < 0.05).Conclusion: The results are consistent with a partial restoration of the blood-brain barrier following treatment.Further, treatment with the selected liposomal drugs gave rise to blood vessels that were morphologically moremature and a vascular network that was more evenly distributed. Taken together the results suggest thattreatment can lead to normalization of GBM blood vessel the structure and function. An in vitro assay designed toassess the effects of extended drug exposure on endothelial cells showed that selective cytotoxic activity againstproliferating endothelial cells could explain the effects of liposomal formulations on the angiogenic tumorvasculature.Keywords: glioblastoma multiforme vasculature normalization, liposomal drugs, endothelial cells* Correspondence: mverreau@bccrc.ca1Experimental Therapeutics, British Columbia Cancer Agency, 675 West10thAvenue, Vancouver, BC V5Z 1L3, CanadaFull list of author information is available at the end of the articleVerreault et al. BMC Cancer 2011, 11:124© 2011 Verreault et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (, which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.BackgroundGlioblastoma (GBM) tumors are largely refractory tosystemic treatments; the median survival time forpatients with GBM is 10 months and the 2-year survivalrate is less than 10%. Chemotherapy for GBM is com-promised in part by the blood-brain barrier limitingdrug access to the malignant cells. In addition, pre-clini-cal models showed that GBM tumors are poorly per-fused [1,2] due to factors such as reduced blood flowrates, elevated hematocrit and interstitial fluid pressure,and an increase in geometric resistance [3-6], all ofwhich impede drug delivery to the tumor tissue. Strate-gies which improve vascular function in GBM tumorsshould improve the delivery of other drugs capable ofcrossing the blood brain barrier and this should be asso-ciated with an increase in therapeutic activity.Our laboratory has previously characterized anddescribed the effects of a liposomal formulation of irino-tecan (Irinophore C™) [7,8]. Encapsulation of irinotecaninto liposomes improved the pharmacokinetic profile ofthe drug and its active metabolite, SN-38. More specifi-cally, administration of Irinophore C™ resulted in a1000-fold increase in the area-under-the-curve ofplasma irinotecan concentration when compared to freedrug (Camptosar). In addition, following irinophore C™injection, the plasma levels of SN-38 were maintained atconcentrations that were up to 40-fold higher than thatachieved following injection of free drug [7]. Followingirinophore C™ treatment, the s.c. (subcutaneous) color-ectal tumors (HT-29) exhibited more functional tumorblood vessels, reduced hypoxia, and increased tumorperfusion. Importantly, these changes in tumor vascula-ture were associated with increased tumor uptake ofdoxorubicin and 5-FU given intravenously [8]. The latterdata were consistent with the idea that the tumor vascu-lature in the treated tumors acquires a more “normal-like” function; an effect of anti-angiogenic therapiesdescribed as ‘normalization’ [9,10].The primary goal of the studies reported here was todetermine whether Irinophore C™ is efficacious inmodels of GBM, and whether treatment with this drugformulation would also result in normalization of GBMvasculature. The effects of Irinophore C™ on thegrowth rates and vascular function of the HT-29 color-ectal cancer model was attributed to significantincreases in the drug circulation lifetime and plasmaconcentration when encapsulated in liposomes [7,8]. Wefurther reasoned that liposomal formulations of otherdrugs with known activity against proliferating endothe-lial cells should have preferential cytotoxicity towardsangiogenic tumor vessels and could potentially also ‘nor-malize’ the chaotic and erratic vasculature of tumors.Thus, part of these studies assessed the effects of liposo-mal vincristine [11] and doxorubicin (Caelyx®) ontumor vasculature. Vincristine has previously beenshown to be active against proliferating endothelial cells[12]. Liposomal formulations of doxorubicin have alsobeen shown to have direct effects on tumor associatedvasculature [13-15].The data reported here assess the effects of IrinophoreC™, Caelyx® (a commercially available and FDA-approved liposomal formulation of doxorubicin), andliposomal vincristine on tumor vasculature in subcuta-neous and orthotopic models of GBM. The results indi-cate that Irinophore C™ was the most activeformulation when using treatment endpoints based onchanges in tumor size as well as tumor vascular mor-phology and function in GBM grown subcutaneouslyand orthotopically. The effects were consistent with theidea that following treatment, there was normalizationof tumor vasculature. In the subcutaneous tumors, vas-cular ‘normalization’ was associated with increasedtumor uptake of Hoechst 33342, while in the orthotopicglioma tumors, treatment-induced vascular ‘normaliza-tion’ was associated with decreased tumor uptake ofHoechst 33342.MethodsCell cultureAdult dermal human microvascular endothelial cells (d-HMVEC; Cambrex Bio Science, Walkersville, MD),Human brain microvascular endothelial cells (HBMEC;ScienCell Research Laboratories, San Diego, California)and U251MG glioblastoma cells (American Type Cul-ture Collection, Manassas, VA) were characterized andauthenticated by the cell banks using immunofluores-cent methods and used for a maximum of eight passagesfor the endothelial cells and fifteen passages forU251MG. Stock cells lines were maintained in theabsence of penicillin and streptomycin and screened formycoplasma prior to preparing a stock of cells that werefrozen for use in experiments. D-HMVEC cells weremaintained in Endothelial Cell Basal Medium-2 (Clo-netics®, Lonza, Basel, Switzerland) supplemented with 5ng/mL Fibroblast Growth Factor, 20 ng/mL VascularEndothelial Growth Factor, 10 ng/mL Epidermal GrowthFactor (Clonetics®, Lonza), 10 unit/mL Heparin (Phar-maceutical Partners of Canada) 1% L-glutamine, 1%penicillin/streptomycin (Stem Cell Technologies, Van-couver, BC, Canada) and 10% Fetal Bovine Serum (FBS;Hyclone, Logan, UT), and plated in 1% gelatin (Sigma,Oakville, ON, Canada) pre-coated dish. HBMEC cellswere maintained in Endothelial Cell Medium supple-mented with Endothelial Cell Growth Supplement(ScienCell Research Laboratories) containing 5 μg/mLInsulin, 10 ng/mL Epidermal Growth Factor, 2 ng/mLFibroblast Growth Factor, 2 ng/mL Insulin-like GrowthFactor-1, 2 ng/mL Vascular Endothelial Growth Factor,Verreault et al. BMC Cancer 2011, 11:124 2 of 181 μg/mL hydrocortisone, 5% FBS and 1% penicillin/streptomycin, and plated in 15 μg/mL fibronectin(Sigma) pre-coated dish. U251MG cells were maintainedin DMEM medium supplemented with 1% L-glutamine,1% penicillin/streptomycin (Stem Cell Technologies,Vancouver, BC, Canada) and 10% FBS (Hyclone, Logan,UT). All cell lines were cultured at 37°C in a humidifiedatmosphere containing 5% CO2, and used during expo-nential growth phase unless otherwise stated.GBM animal model s.c. and orthotopicAll protocols involving work with live animals werereviewed and approved by the University of BritishColumbia Animal Care Committee (certificate ofapproval # A07-0423). For the subcutaneous GBMmodel, U251MG cells (5 × 106) were implanted subcuta-neously into the backs of Rag2M mice (7-10 weeks oldfemales, n = 9). To generate orthotopic GBM tumors,U251MG (7.5 × 104) cells were implanted into the rightcaudate nucleus-putamen (ML -1.5 mm; AP +1 mm;DV -3.5 mm) of mice (n = 5-6) using a stereotaxic injec-tion frame (Stoelting Company, Wood Dale, IL). Ani-mals were treated with 25 mg/kg Irinophore C™, 2 mg/kg liposomal vincristine or 15 mg/kg doxorubicin lipo-some (Caelyx®, Schering-Plough, QC, Canada) i.v. onday 21, 28 and 35 after inoculation. Dosing of liposomalvincristine and Caelyx® resulted in less than 5% bodyweight loss, while Irinophore C™ treatment did notcause any change in body weight. Previous tests in ourlaboratory have shown that the maximum tolerated sin-gle doses for Irinophore C™, Caelyx® and liposomalvincristine are >120 mg/kg, 17 mg/kg and 3 mg/kg,respectively. Irinophore C™ [16] and liposomal vincris-tine [17] were prepared as described previously. S.c.tumor size was measured throughout the study by cali-per and tumor weights were extrapolated from the mea-surements using the following formula: mg = (tumorwidth^2 × tumor length)/2 [18]. Mice were injectedwith Hoechst 33342 (1.2 mg/mouse; Sigma) twelve (s.c.model) or twenty (orthotopic model) minutes prior tosacrifice on day 42. This timing was chosen based onprevious study [8] and tests (not shown) aimed at deter-mining the optimal timing for Hoechst 33342 injectionwithout saturation of the tissue and before any decreasein Hoechst 33342 staining could be observed due topossible metabolic elimination. All animals were termi-nated by CO2 asphyxiation and s.c. tumors or brainswere harvested and cryopreserved in OCT (Sakura Fine-tek, CA) on dry ice and stored at -80°C.Hoechst 33342, Ki67, CD31, VEGFR2, EF5, Collagen IV,NG2 and nuclei density staining and quantificationOptimal Cutting Temperature compound (OCT)-pre-served s.c. tumors were cryosectioned using a LeicaCM1850 Cryostat (Leica, ON, Canada) and 10 μm sec-tions were collected in the middle of each tumor. OCTpreserved brains were cryosectioned and 10 μm sectionswere collected from the Bregma +1.0 location. Sectionswere fixed in a 1:1 mixture of acetone:methanol for 15minutes at room temperature, then blocked with block-ing buffer (Odyssey blocking buffer, Rockland, PA) for 1hour at room temperature. Sections were stained with ratanti-mouse CD31 antibody (1:100 dilution, PharMingen#550274, BD Biosciences), rabbit anti-human Ki-67 (Invi-trogen #18-0191z; 1:100), rabbit anti-human/mouse vas-cular endothelial growth factor receptor 2 antibody(VEGFR2; 1:100; Cell Signaling technology #2479, NEB,Pickering, ON, Canada), rabbit anti-Collagen IV antibody(1:400, Abcam # ab19808, Cambridge, MA) and mouseanti-NG2 chondroitin sulfate proteoglycan antibody(1:100, Millipore # MAB5384, Billerica, MA). Primaryantibodies were incubated on sections overnight at 4°C.Secondary antibodies (Alexa 488 goat anti-rat #A11006,Alexa 546 goat anti-rabbit #A-11035 and Alexa 633 goatanti-mouse #A-21126, 1:200, Invitrogen) were incubatedfor 1 hr at room temperature. Nuclei were stained withDraq5 (Biostatus, Leicestershire, UK; 1:200) for 30 min at37°C. Slides were mounted with PBS and imaged forAlexa 488 (L5 filter), Hoechst 33342 (A4 filter), Alexa546 (Cy3 filter), Cy5 (Cy5 filter) and Draq5 (Cy5 filter)using a robotic fluorescence microscope (LeicaDM6000B, Leica, ON, Canada) and a composite colorimage of these markers was produced (Surveyor software,Objective Imaging Ltd.). Thresholds for each markerwere set using Photoshop; the threshold level was setusing a scale from 1 to 255 units, and was defined at 2units higher than the minimal level necessary to obtain anegative signal for non-specific staining, and was kept thesame for all sections. Acquired images were quantifiedfor positive pixels or colocalization (double-positive pix-els) using an in-house segmentation algorithm, normal-ized to the number of pixels in the tumor area andexpressed as positive fraction (positive pixels divided bynon-necrotic tumor area; MATLAB, The Mathworks,Natick, MA). Non-necrotic tumor areas were defined bycropping out necrotic and non-tumor tissue on the basisof positive Ki-67 and Draq5 co-stained sections and werequantified using the same in-house algorithm. Colocali-zation was considered positive when two positive pixelsfrom one stain of interest were located within a 3 pixelsradius from one pixel of the other stain of interest. Ofnote, one cell nucleus measures between 3 and 6 pixels.Blood vessel diameter was defined by taking 10 measure-ments/tumor section in a 15 × 15 cm box at 200% mag-nification using Photoshop, and was expressed in pixels.For differential analysis between the tumor’s center andperiphery, the boundary between the tumor center andperiphery area was established at 20% of tumor diameterVerreault et al. BMC Cancer 2011, 11:124 3 of 18distance from tumor margin. Another set of sections wasstained with hematoxylin and eosin for histopathologyanalysis. The fraction of collagen IV-free blood vesselswas defined as Collagen IV negative/CD31 positive pixelsover total CD31 pixels. The fraction of NG2-free bloodvessels was defined as NG2 negative/CD31 positive pixelsover total CD31 pixels. The amount of basement mem-brane empty sleeves was defined as CD31 negative pix-els/collagen IV positive pixels divided by the total non-necrotic tumor area.Magnetic Resonance Imaging and Ktrans measurement inU251MG orthotopic tumorsAll magnetic resonance experiments were carried outusing a 7.0 Tesla MR scanner (Bruker, Ettlingen Ger-many). A Bruker (Ettlingen, Germany) volume coil (innerdiameter of 7 cm) and rectangular surface coil (1.7 × 1.4cm) was used for signal transmission and receptionrespectively. The coil was tuned to the hydrogen protonfrequency (300.3 MHz). The Ktrans values were obtainedfrom serial images acquired to monitor changes in theconcentration of a MR-visible contrast agent (GD-DTPA;Bayer Schering Pharma) within each pixel, during theinitial uptake and subsequent washout of the agent in thetumor. The MRI scans follow the protocol reported byLyng et al. [19]; briefly, mice were anaesthetized with iso-fluorane (5% induction, 2% maintenance), a catheterinserted into the lateral tail vein and the animal was placedsupine with its head above the surface coil. A proton-den-sity weighted scan was first acquired to serve as a baselinefor conversion of pixel intensity to absolute concentrationvalues of the contrast agent. A volume equivalent to 10 uLper gram body weight of the contrast agent (0.03 M Gd-DTPA in saline) was injected via the tail vein catheter in aperiod of 10-15 seconds. The contrast series consisted of a3D RF-spoiled Fast Low Angle Shot (FLASH) sequencewith timing and resolution parameters as follows: echotime/repetition time = 2.8/9.2 ms, Field of view = 1.92 ×1.92 × 1.6 cm, Matrix size = 128 × 128 × 16 cm, acquisi-tion time per image = 9.45 seconds. Twenty baseline scanswere acquired before contrast agent injection and 250scans were acquired afterwards, resulting in a total acquisi-tion time of 43 minutes. The concentration-time curve foreach pixel was fit to a two-compartment Kety model [20]which describes the pharmacokinetics of the contrastagent using three parameters: ve (volume of extracellularextravascular space), Ktrans (volume transfer constantbetween the vasculature and tissue compartment) and Vp(fractional volume of the vascular compartment).In vitro endothelial cell exposure and nuclei countFor proliferative conditions, Dermal Human MicroVas-cular Endothelial Cells (d-HMVEC; 600 cells/well) andHuman Brain Microvascular Endothelial Cells(HBMEC; 5000 cells/well) were plated in black 96-wellplates (Optilux™, BD Biosciences, Mississauga, ON,Canada) and drugs were added the day after. For non-proliferative conditions, d-HMVEC cells (5000 cells/well) and HBMEC (50000 cells/well) were plated inblack 96 well plates and drugs were added four daysafter. Irinotecan (Sandoz, QC, Canada), SN-38 (LKTLaboratories, MN, USA), vincristine (Novopharm, ON,Canada), docetaxel, paclitaxel (Taxol®; Bristol MyersSquibb Canada, QC, Canada) and doxorubicin (Adria-mycinTM/MC, Pfizer, QC, Canada) were added in con-centrations ranging from 1-100,000 picoMolar on cellsand replaced daily for 7 days. At the end of drug treat-ment, cells were fixed with 3.5% paraformaldehyde(Electron Microscopy Sciences, PA) for 15 minutes at-20°C, permeabilized with 0.1% Triton (Perkin-Elmer,MA) in PBS for 10 minutes at room temperature,blocked for 1 hr at 4°C (Odyssey blocking buffer,Rockland, PA) and incubated overnight with Ki67 anti-body (Invitrogen #18-0191z; 1:100 dilution in blockingbuffer). Cells were then incubated with Anti-rabbitAlexa 488 secondary antibody (Molecular Probe#A11034, Invitrogen; 1:200 in blocking buffer) for 1 hrat room temperature. Nuclei were stained with Draq5dye (Biostatus, Leicestershire, UK; 1:200 in PBS) for 30min at 37°C. Twenty fluorescent photographs/well(Alexa 488 emission: 475 nm, excitation: 535 nm;Draq5 emission: 620 nm, excitation: 700 nm) weretaken at 10 × magnification using an InCell Analyzer1000 (Amersham Bioscience) and the total nucleicount (Draq5 stained nuclei) as well as Ki67 expressingnuclei count (Draq5 and Alexa 488 double stainednuclei) were quantified using InCell Developer Tool-box software (Amersham Bioscience, GE Healthcare,Baie d’Urfe, QC, Canada). Dose-response curves gener-ated from total nuclei count were used to calculatedrug concentrations causing a decrease in endothelialcell nuclei count by 20% (fraction affected: Fa = 0.2),50% (Fa = 0.5), 75% (Fa = 0.75) and 90% (Fa = 0.9)and compared for both proliferative and non-prolifera-tive cells. All data points represent the average of 3independent experiments in triplicate +/- S.E.M.Statistical analysisAll statistical data was collected using GraphPad Prism(San Diego, CA). Because all treatment drugs were chosenbased on previous rationale justifying their inclusion in thestudy, the experimental design should not be regarded as ascreening assay and statistical analysis was done using thesingle comparison non-parametric two-tailed Mann Whit-ney test and no correction was made for multiple compar-isons. All data are expressed +/- S.E.M.Verreault et al. BMC Cancer 2011, 11:124 4 of 18ResultsIrinophore C™, Caelyx® and liposomal vincristine inhibittumor growth and increase Hoechst 33342 delivery insubcutaneous GBM tumorsRag2M mice bearing s.c. U251MG tumors (n = 9) weretreated i.v. weekly for 3 weeks with 25 mg/kg IrinophoreC™, 15 mg/kg Caelyx® and 2 mg/kg liposomal vincris-tine. Tumor growth was monitored during the entiretreatment period, and tumors were harvested 7 daysafter the last treatment. As noted in Figure 1a, the threedrugs inhibited tumor growth significantly compared tountreated control (p < 0.05-0.001). At the end of thestudy (day 42), the weight of treated tumors rangedfrom 34 to 80 mg compared to an average of 502 mgfor untreated control animals. A representative tumorsection (H&E) derived from each treatment group isalso provided in Figure 1a. The total non-necrotictumor area (excluding necrotic and non-tumor area)measured in number of image pixels for each treatedgroup is summarized in Figure 1b. The measurementsof area of viable tumor tissue correlated with the tumorweight measurement and was significantly reduced forall treatment groups (compared to untreated tumors; p< 0.0001). The proliferation marker Ki67 was used toestimate the fraction of viable cells undergoing activeproliferation within the tumor (positive Ki67 stainingdivided by total viable tissue, expressed as Ki67 positivefraction). Liposomal vincristine had no apparent effecton the Ki67 staining compared to control tumors.Treatment with Irinophore C™ caused a 2-fold decreasein Ki67 staining (p < 0.01). In contrast, a significant (p <0.01) increase in Ki67 staining was observed in tumorsfrom animals treated with Caelyx® (Figure 1b). It shouldbe noted that Caelyx® treatment was also associatedwith enlarged tumor cell nuclei (see arrow heads ininsert H&E image Figure 1a) and this may suggest thatthe treatment promoted cell cycle arrest [21]. Thisobservation is in accordance with previously publishedfindings on the effects of doxorubicin on cell cycle[22-24] and the fact that cellular Ki67 antigen has beenshown to accumulate in some types of cell cycle arrest[25]. Finally, a decrease in number of cell nuclei perarea (nuclei density) with a concomitant increase inconnective tissue was observed by examination of theH&E stained sections in tumors from mice treated withIrinophore C™.The effects of the selected liposomal drugs on tumorblood vessels were also evaluated. As summarized inFigure 1c, the CD31 staining (positive CD31 fraction)did not change significantly when comparing tumorsfrom control animals to those from treated animals.Prior to sacrifice, animals were injected with Hoechst33342, a marker for tumor perfusion that was previouslyvalidated by correlation with Ktrans measurements [8].Total Hoechst 33342 staining in viable tissue (positiveHoechst 33342 fraction) was increased in the tumorsobtained from treated animals (p < 0.01-0.001; Figure 1c).CD31 and Hoechst 33342 co-staining was measured toprovide an indication of changes in functional bloodvessels [8]. The results, summarized in Figure 1c, indicatethat the number of functional blood vessels increasedsignificantly (p < 0.05) in Caelyx® treated tumors whilethere were no significant changes observed in tumorsfrom Irinophore C™ and liposomal vincristine treatedanimals.Irinophore C™, Caelyx® and liposomal vincristine inhibittumor growth and decrease Hoechst33342 delivery inorthotopic GBM tumorsRag2M mice (n = 5 or 6) were inoculated with U251MGcells orthotopically (see Methods) and 21 days later theanimals were treated i.v. (once weekly for 3 weeks) with25 mg/kg Irinophore C™, 15 mg/kg Caelyx® and 2 mg/kg liposomal vincristine. Forty-two days after cell inocu-lation, animals were sacrificed and their brains har-vested. A representative tissue section (Hematoxylin andEosin; H&E) showing the site of tumor growth (darkblue) within the brain of treated animals is provided foreach treatment group in Figure 2a. Insert images havebeen included to show that following treatments, thetumor nuclear density drops slightly when compared tountreated controls. The average total non-necrotictumor tissue in the tumor area for each treatmentgroup was quantified to provide a measure of efficacy(Figure 2b). There was a significant reduction in tumorarea for all treatment groups when compared to con-trols (p < 0.0001). In contrast to the results obtainedwith the s.c. glioma model, there was no significantchanges in Ki67 staining observed following treatment(Figure 2b).Prior to sacrifice, animals were also injected withHoechst 33342. In tumors from untreated control mice,Hoechst 33342 staining was significantly greater intumor tissue compared to matched regions of normalbrain tissue (0.398 +/- 0.083 and 0.023 +/- 0.015 pixels/unit area, respectively; p < 0.01; data not shown). Thisstaining pattern has been described elsewhere [26,27]and is consistent with the fact that Hoechst 33342 doesnot cross the blood-brain barrier. Interestingly, the datasummarized in Figure 2c show that Hoechst 33342staining in the orthotopic tumor tissue from animalstreated with the liposomal drugs was significantlyreduced (p < 0.01) when compared to tumors fromcontrol animals. The decrease in Hoechst 33342 stainingin orthotopic tumors from treated animals was inmarked contrast to treatment-induced increases inVerreault et al. BMC Cancer 2011, 11:124 5 of 18Hoechst 33342 staining noted for tumors derived fromthe same cell line (U251MG) and grown subcutaneously(Figure 1c).No significant changes in overall CD31 staining (pix-els/unit area) (Figure 2c) were noted in the orthotopictumors obtained from treated animals (compared tocontrols). However, CD31/Hoechst 33342 co-stainingwas significantly reduced (p < 0.01-0.05) in tumorsfrom treated animals when compared to control animals(Figure 2c). Moreover, treatment of orthotopic tumorFigure 1 Irinophore C™, liposomal vincristine and Caelyx® significantly inhibits tumor growth, decreases proliferation and increasestumor perfusion in subcutaneous GBM tumors. a) Representative H&E sections of tumors from each treatment group show the efficacy ofthe treatments in controlling tumor growth. Arrow heads indicate enlarged nuclei associated with Caelyx® treatment. Tumor weights werecalculated on the basis of caliper measurements; arrows indicate the treatment days. The Irinophore C™ statistical significance is indicated bybottom stars, while Caelyx® and liposomal vincristine statistical significances are indicated by top stars (*p-value ≤ 0.05; **p-value ≤ 0.01; ***p-value ≤ 0.001) b) The area of viable tissue in tumor sections following treatment was expressed in number of pixels and correlates well withtumor volumes (■, left axis). The fraction of viable, actively proliferating cells (■, right axis) in the tumors was significantly decreased byIrinophore C™. Ki67 staining was also increased in Caelyx®-treated tumors. c) Hoechst 33342 perfusion in the tumors was increased significantlyby Irinophore C™ and Caelyx® treatment (■, left axis). The number of endothelial cells per unit area of viable tissue was unchanged by thetreatments (■, right axis); however, the fraction of endothelial cells that were perfused (CD31 and Hoechst 33342 positive; □, right axis) wasincreased by treatment with Caelyx®. Statistical significances are indicated (*p-value ≤ 0.05; **p-value ≤ 0.01; ***p-value ≤ 0.001).Verreault et al. BMC Cancer 2011, 11:124 6 of 18bearing animals with Irinophore C™ was associatedwith a significant (p < 0.05) increase in CD31 stainingin the center of tumors when compared to untreatedtumors (Figure 2d; p < 0.05).Assessing vascular normalization in GBM tumors fromanimals treated with Irinophore C™, Caelyx® or liposomalvincristineSeveral structural determinants, described as indicatorsof vascular normalization [28-30], were assessed in theorthotopic and s.c. GBM tumor models following treat-ment and these data were compared to tumors fromuntreated control animals. The parameters evaluatedincluded: (i) the extent of discontinuous basement mem-brane (collagen IV-free CD31 pixels) in the tumor tis-sue, (ii) the fraction of pericyte-uncovered blood vessels(NG2-free CD31 pixels) in the tumor tissue and (iii) theblood vessel diameter. Furthermore, the proportion ofempty basement membrane sleeves (CD31-free collagenIV pixels) was evaluated as an indication of regressionof pre-existing blood vessels [9]. Treatment-inducedchanges in these factors are summarized in Figures 3(s.c. tumors) and 4 (orthotopic tumors).In s.c. GBM tumors, the fraction of NG2-free bloodvessels was reduced by 25% in tumors from animalstreated with Irinophore C™ (p < 0.05; Figure 3a).Decreases in NG2-free blood vessels were also noted intumors from animals treated with Caelyx® (p = 0.071)or liposomal vincristine (p = 0.121); but the effects werenot considered significant. The number of collagen IV-free blood vessels was decreased in s.c. tumors from ani-mals treated with Irinophore C™ or Caelyx® (41-75%Figure 2 Orthotopic GBM tumors treated with Irinophore C™, liposomal vincristine and Caelyx® are significantly smaller thanuntreated controls. a) Representative H&E images of brain sections from mice in each treatment group show that the area of tumor tissue(dark blue) from treated animals are smaller than untreated controls. b) Tumor areas were quantified in number of pixels, and used as a measureof treatment induced reduction of the tumor mass (■, left axis). No significant changes in proliferative activity (■, right axis) were observed. c)Hoechst 33342 staining was reduced significantly following treatments with the three liposomal treatments (■, left axis). The total number ofendothelial cells per unit area of viable tissue (■, right axis) was unchanged across all groups, but the fraction of endothelial cells that were co-stained with Hoechst 33342 was significantly reduced (□, right axis). d) The density of endothelial cells (positive CD31 pixels divided by peripheryor center tumor area pixels) in the center of tumors treated with Irinophore C™ was significantly higher compared to control tumors (□). Nochanges in endothelial cell density were seen in the total tumor area (■) or the periphery of tumors (■). Statistical significances are indicated (*p-value ≤ 0.05; **p-value ≤ 0.01; ***p-value ≤ 0.001). Non-significant trends are indicated (&p-value = 0.067).Verreault et al. BMC Cancer 2011, 11:124 7 of 18decrease; p < 0.05-0.001; Figure 3a). Blood vessel dia-meter was also reduced (32%-51%; p < 0.001) in s.c.tumors from all treatments groups. Finally, the numberof empty basement membrane sleeves in tumors fromIrinophore C™ and Caelyx® treated animals wasincreased 3.4- to 3.8-fold following treatment (p <0.0001). A similar effect was noted for tumors fromanimals treated with liposomal vincristine, but theeffect was not considered significant (p = 0.054).Representative immunofluorescence micrographs high-lighting the effects of Irinophore C™ treatment on thetumor vasculature of s.c. U251MG tumors (Figure 3c)compared to untreated tumor (Figure 3b) are providedto support the results summarized in Figure 3a.Similar results were obtained when evaluating theorthotopic U251MG tumors from treated animalscompared to controls. In addition, histological assess-ments of brain tissue surrounding the tumor allowedFigure 3 Irinophore C™, liposomal vincristine and Caelyx® treatments are associated with vascular normalization of the tumorvasculature in subcutaneous GBM tumors. a) The diameters of tumor blood vessels were reduced significantly by all three treatmentscompared to control tumors (■, left axis). The fraction of NG2-free CD31 pixels (■, right axis), collagen IV-free CD31 pixels (■, right axis) werereduced in Irinophore C™ treated tumors indicating that fewer immature vessels are present following treatments. The proportion of emptybasement membranes (CD31 free-collagen IV staining; □, right axis) in the viable tissue was also reduced by all liposomal treatments. b and c)Representative and merged images of CD31, Collagen IV and NG2 staining for control tumors (b) and tumors from Irinophore C™ treated mice(c). A reduction in blood vessel diameter (CD31; green) and an increase in basement membrane coverage of blood vessels (collagen IV; yellow)following treatment can be seen. Following Irinophore C™ treatment, more pericytes are present (NG2; red) and more endothelial cells areassociated with the pericytes (merged image, green and red). Treatment with Irinophore C™ also results in an increase in empty basementmembrane sleeves (i.e. CD31 free-collagen IV; yellow). The entire image represents non-necrotic and viable tissue. Statistical significances areindicated (*p-value ≤ 0.05; **p-value ≤ 0.01; ***p-value ≤ 0.001; ****p-value ≤ 0.0001). Non-significant trends are indicated (&p-value = 0.121;&&p-value = 0.071; &&&p-value = 0.054).Verreault et al. BMC Cancer 2011, 11:124 8 of 18comparisons between vessels in the tumor vs. normalbrain tissue. The fraction of collagen IV-free blood ves-sels in normal brain (0.049 +/- 0.015) was 69% lower (p< 0.05) than that observed in tumor tissue from controluntreated animals (0.160 +/- 0.033), indicating that theorganization of the basement membrane architecture isdecreased in the tumor compared to normal tissue (datanot shown). Tumors from animals treated with Irino-phore C™ showed a significant 71% (p < 0.05) decreasein the fraction of collagen IV-free blood vessels whencompared to tumors from control animals (Figure 4a).A similar effect was observed in tumors from animalstreated with Caelyx®, but the effect was not consideredsignificant (p = 0.064). In normal brain tissue, bloodvessel diameters were 54% smaller (4.9 +/- 0.5 pixels; p< 0.0011) than blood vessel diameters observed inorthotopic tumor tissue obtained from untreated ani-mals (10.9 +/- 0.6 pixels; data not shown). Orthotopictumors from animals treated with Irinophore C™ orCaelyx® exhibited a reduction in blood vessels diametersof 39% (p < 0.01; Figure 4a) when compared to controltumors. In contrast to results obtained with the s.c.tumors of treated animals, the level of empty basementmembrane sleeves (Collagen IV-free CD31 staining) inthe orthotopic tumors did not change following treat-ment (Figure 4a). It should be noted that the level ofempty basement membrane sleeves in the normal braintissue (0.035 +/- 0.009) was found to be similar to thatmeasured in orthotopic tumor tissue from untreatedanimals (0.047 +/- 0.009) (data not shown). Treatmentsdid not induce significant changes in fraction of NG2-free blood vessels (Figure 4a). The fraction of NG2-freevessels in the normal brain could not be evaluated asNG2 proteoglycan was found at the surface of polyden-drocytes, a subpopulation of glial cells found in thebrain [31]. Representative immunofluorescence micro-graphs illustrating the effects of Irinophore C™ treat-ment on the orthotopic tumor vasculature are providedin Figure 4b. Normal brain tissue sections are shown inFigure 4c for comparison.Magnetic resonance imaging-measured changes invascular permeability/flow (Ktrans)The results summarized thus far are consistent with theidea that following treatment of animals bearing GBMtumors with lipid-based nanopharmaceutical formula-tions of vincristine, doxorubicin and irinotecan, there isa “normalization” of blood vessel structure. When con-sidering these effects along with the antitumor activity,the greatest effects were observed following treatmentwith Irinophore C™. In order to confirm the idea of aIrinophore C™-induced vascular normalization, non-invasive magnetic resonance imaging was used to assessKtrans, a volume transfer constant of a solute betweenthe blood vessels and extra-cellular tissue compartment,in orthotopic tumors grown in untreated and IrinophoreC™treated mice. The median values of Ktrans for thetumors within the control and treated groups have beensummarized Figure 5. The results demonstrate that themedian Ktrans value in untreated tumors was ~7 timesgreater than in treated tumors (0.0232 and 0.0034 ml/g/min, respectively, p < 0.05). It should be noted that thevalues for Ktrans in tumors from untreated animals weremore variable when compared to the tumors from trea-ted mice (s.e.m ± 0.010 and ± 0.0003, respectively).In vitro studies on endothelial cells mimicking theextended drug exposure achieved when using liposomaldrug delivery formulationsIn an attempt to better understand the effects of liposo-mal formulations used here on tumor vasculature, an invitro endothelial cell assay was used to assess the impactof extended drug exposure. It is well established thatthese liposomal formulations engender significantincreases in plasma drug concentrations over extendedtime periods following intravenous administration [7,11].Thus, an extended drug exposure protocol was used toassess the effects of drugs in a model representative ofthe endothelial cells forming vessels in the subcutaneousor brain microenvironment. Dermal Human MicroVas-cular Endothelial Cells (d-HMVEC) and Human BrainMicrovascular Endothelial Cells (HBMEC) were culturedunder proliferative or non-proliferative conditions andexposed to the indicated drugs for 7 days. As illustratedin Figure 6a, the total nuclei count and the number ofnuclei expressing the Ki67 proliferation marker werequantified using high content screening (Incell analyzer1000) to discriminate between cytotoxic (reduction intotal number of nuclei) and cell proliferation inhibitoryeffects (reduction in Ki67 expressing fraction). Underproliferative conditions, the nuclei count for the endothe-lial cell lines used increased up to 3-fold over the 7 daytime period. The Ki67 expressing nuclei fraction rangedfrom 42 to 68% over this time frame (Figure 6a and 6b).Under non-proliferative conditions, the nuclei count forcell lines (d-HMVEC and HBMEC) remained unchangedfrom day 1 to day 7, and the Ki67 expressing nuclei frac-tion ranged from 7 to 31% (Figure 6a and 6b).The activity of the drugs against the cells maintainedunder the two conditions was compared to assess theselectivity of the drugs for proliferating endothelial cellscompared to non-proliferating endothelial cells. For alldrugs used in this study, the dose-response curves forKi67 expressing nuclei of proliferating cells matchedthe ones for the total nuclei count, suggestingthat the drugs tested were cytotoxic rather thanVerreault et al. BMC Cancer 2011, 11:124 9 of 18Figure 4 Irinophore C™, liposomal vincristine and Caelyx® treatments are associated with vascular normalization of the tumorvasculature in orthotopic GBM tumors. a) Vessel diameters (■, left axis) and the fraction of collagen IV-free CD31 pixels (■, right axis) inorthotopic GBM tumors were reduced by Irinophore C™ and Caelyx®. However, no changes were seen in the fraction of NG2-free CD31 positiveendothelial cells (■, right axis) or Collagen IV-free CD31 positive endothelial cells (□, right axis). b) Representative images from untreated andIrinophore C™ treated tumors; similar images for normal brain tissue are shown for comparison. (c) Blood vessel diameters (CD31; green) arereduced by Irinophore C™ treatment. The basement membrane (collagen IV; yellow) is partially restored by treatment with Irinophore C™. Theentire image represents non-necrotic and viable tissue. Statistical significances are indicated (*p-value ≤ 0.05; **p-value ≤ 0.01). Non-significanttrends are indicated (&p-value = 0.064).Verreault et al. BMC Cancer 2011, 11:124 10 of 18anti-proliferative. Representative dose-response curvesfor d-HMVECs and HBMECs exposed to SN-38, theactive metabolite of irinotecan, under proliferative andnon-proliferative conditions are shown in Figure 6b.The data indicates that SN-38 is significantly moreactive against proliferating endothelial cells then non-proliferating cells. In an effort to highlight differences indrug activity under proliferating and non-proliferatingconditions, drug concentrations decreasing d-HMVECtotal nuclei count by 20% (fraction affected: Fa = 0.2),50% (Fa = 0.5), 75% (Fa = 0.75) and 90% (Fa = 0.9) werecalculated from the dose-response curves and compa-red for both proliferative and non-proliferative cells(Figure 7). For example, results obtained at Fa = 0.75indicate that the greatest differential effects wereobserved when using SN-38 and vincristine, where thedrug dose required to achieve a 75% decrease in nucleicount under proliferation conditions were at least 100-and 90.9-times lower, respectively, than the drug doserequired to achieve the same effect level under non-pro-liferative conditions. These effects were much greaterthan those seen using the positive control compoundsdocetaxel and paclitaxel. In contrast, there was little orno difference in the concentrations of irinotecan or dox-orubicin required to achieve a Fa of 0.75 under prolifer-ating and non-proliferating conditions. Similar resultswere obtained when using HBMECs (Figure 8). Itshould be noted that the drug doses required to achievea Fa value of 0.5 for SN-38 and vincristine was 45 to5000 times greater for U251MG glioblastoma cells whencompared to the proliferating endothelial cells (data notshown) and the increased specificity for proliferatingendothelial cells has been noted previously for paclitaxeland SN-38 when compared against human colorectaland breast cancer cells [32,33].DiscussionStudies over the last few decades have established thatliposomal formulations of selected antineoplastic agentscan be more effective than the same drug administeredin free form. Liposomal formulations of anticancerdrugs are known to have long circulation half-lives invivo, and release the drug slowly over time [7,11]. Thus,the pharmacological properties of a drug given in itsfree form (e.g. via bolus injection or slow infusion) ischanged dramatically by encapsulation in liposome. As aresult, one might anticipate that the use of liposomaldrugs will expose tumors to drugs for extended periodsof time when compared to treatment with the free drug.This, of course, is well established in the literature andhas been explained on the basis of the enhanced perme-ability and retention effect known to promote accumula-tion of intravenously administered liposomal drugformulations in tumors [34]. What is often not consid-ered in studies with liposomal formulations is that theseformulations constantly release the associated drugwhile in the circulation compartment, thereby extendingthe presence of the drug in the plasma compartment.This study tries to address whether part of the treat-ment benefits could be attributed to direct effects of thefree drug (available in the blood compartment) ontumor vascular endothelial cells. The fact that thesedrug formulations are active against proliferating vascu-lature was anticipated, but not demonstrated to date.Liposomal drug formulations are known to accumulateand release drugs in close proximity to tumor blood ves-sels [14,15]. More intriguing, however, is the possibilitythat exposing the tumor vasculature to low concentra-tions of drug for extended periods may produce effectsthat are comparable to the vascular normalization effectsdescribed in the context of anti-angiogenic therapy[9,10] as discussed below.In the present study, it is demonstrated that Irino-phore C™, Caelyx® and liposomal vincristine are effec-tive against GBM grown subcutaneously ororthotopically (in the brain). The tumor masses in trea-ted animals were significantly smaller compared to con-trol (p < 0.001; Figure 1a), indicating that the liposomaldrugs used in this study are potent against GBM,regardless of the site of tumor growth. Analysis of thetumor tissue, and in particular the vascular morphology,also indicates that treatments affected the tumor vascu-lature to various degrees. Overall, Irinophore C™impacted the vasculature to a greater extent than theother formulations, and generated tumors with bloodvessels that were morphologically more mature. In thesubcutaneous model, Irinophore C™ restored the base-ment membrane architecture, increased the pericytecoverage and reduced blood vessel diameters. The dataFigure 5 Irinophore C™ reduced Ktrans measures compared tovalues obtained from control orthotopic tumors. Individual andmedian (thick line) Ktrans values for untreated and Irinophore C™treated orthotopic tumors with standard error of the mean (thinline). Statistical significance is indicated (*p-value ≤ 0.05).Verreault et al. BMC Cancer 2011, 11:124 11 of 18suggest a restoration of the vessel architecture to a morenormal state. In the more clinically relevant orthotopicmodel, Irinophore C™ treatment restored the basementmembrane architecture and reduced blood vessel dia-meters of the tumor vasculature, again suggesting arestoration of the vessel architecture to a more normalstate. Irinophore C™ treatment also increased the quan-tity of vessel staining in the center of tumors, suggestinga more homogenous distribution of blood across theentire tumor. Further, Irinophore C™ reduced KtransFigure 6 d-HMVEC and HBMEC were plated under proliferative conditions or non-proliferative conditions. a) Representative compositecolor images of d-HMVEC cells are shown; Draq (blue; nuclei), and Ki67 (green). Under proliferative conditions, the number of nuclei and Ki67positive staining are similar; whereas under non-proliferative conditions, the number of nuclei with positive Ki67 staining is much lower. b) Totalnuclei count and Ki67 expressing nuclei fraction of untreated cells for both cell lines under proliferative and non-proliferative conditions on day1 and day 7 after plating (3 independent experiments; 3-21 replicates per experiment). Cells were exposed for 144hrs to increasing drugconcentrations (1-100,000 picoM). Total nuclei count as well as nuclei expressing Ki67 expressing counts were normalized to counts obtainedfrom control untreated cells. Representative data for d-HMVEC and HBMEC exposed to SN-38 is shown (3 independent experiments; 3 replicatesper experiment +/- SEM). Statistical significance is indicated (*p-value < 0.0001) between total nuclei count of proliferative and non-proliferativecells.Verreault et al. BMC Cancer 2011, 11:124 12 of 18values calculated from Dynamic Contrast Enhanced(DCE)-MRI studies significantly. Based on changes invessel morphological appearance, the drop in Ktransvalues was interpreted as a decrease in vessel permeabil-ity [35], and is consistent with the suggestion that Irino-phore C™ treatment improved vascular function in thetumor. The larger variability in Ktrans values determinedin tumors from control animals reflects the random nat-ure of chaotic and leaky blood vessels in individualtumors [36]. It had already been established in s.ctumors that Hoechst 33342 could be used as a markerfor tumor vessel function by validation with Ktrans mea-surements [8], but this had not been done for the ortho-topic GBM tumor described here. It is shown here thatthe observed reduction in Hoechst 33342 staining aftertreatment while total CD31 staining remained constantcorrelates with a reduction in Ktrans measures. Takentogether, these observations strongly suggest animprovement in vascular function. The tumor bloodvessels in tumors from animals treated with IrinophoreC™ behave more like vessels in the normal brain wherethe blood-brain barrier is intact.The concept of ‘blood vessel normalization’ was firstpostulated in the 70s [37] and more recently, the clinicalpotential of vascular normalization has been described[9,10]. As with most solid tumors, the microvasculatureof gliomas is characterized by tortuous and fenestratedvessels with diameters that are larger than normal [38]and discontinuous basement membrane which rarelyencloses pericytes [39]. In glioma [28,29,40], antiangio-genic therapies can stop the growth of tumor vessels,prune immature and inefficient tumor vessels and nor-malize surviving vasculature by increasing the fractionof pericyte-covered vessels, restoring the abnormallythick and irregular basement membrane and reducingthe high vascular permeability of these vessels [9,10]. InFigure 7 Proliferating HMVEC cells are more sensitive to SN-38, docetaxel, paclitaxel, doxorubicin and vincristine than non-proliferating cells. Concentrations at which a Fa of 0.2-0.9 was observed in d-HMVEC total nuclei count for both proliferative and non-proliferative conditions. The fold difference in drug concentration required to achieve the specified Fa is indicated above each pair of columns.Verreault et al. BMC Cancer 2011, 11:124 13 of 18glioblastoma patients, a “vascular normalization index”was defined by changes in vascular permeability (Ktransvalues), microvessel volume and circulating collagen IV.It was found that this index was closely associated withoverall survival and progression-free survival in responseto Cediranib, a pan-VEGFR inhibitor [40]. Pre-clinically,the delivery of temozolomide in an intracerebral modelof glioma increased after treatment with the angiogen-esis inhibitor SU5416. This drug restored capillary archi-tecture and decrease interstitial fluid pressure [41]. Suchstudies offer strong evidence that the tumor vasculaturein GBM is a valid target, and that therapies which ‘nor-malize’ tumor vasculature may improve the delivery of asecond drug at some point in the treatment regimen.The studies described here, together with an earlierpublication [8], offer strong evidence that liposomal for-mulations of selected drugs, and especially IrinophoreC™, induce a normalization of the tumor vasculature.In this study, collagen IV and NG2 were used as mar-kers for basement membrane and pericytes, respectively.However, there is no consensus in the field for a defini-tive marker of these parameters. Other markers used toevaluate basement membranes include nidogen or lami-nin, and desmin or a-smooth muscle actin for pericytes[9,30]. These caveats notwithstanding, the morphologicalchanges observed were associated with changes inHoechst 33342 uptake in the tumor and when using thisparameter, remarkably different results were obtainedFigure 8 Proliferating HBMEC cells are more sensitive to SN-38, docetaxel, paclitaxel, doxorubicin and vincristine than non-proliferating cells. Concentrations at which a Fa of 0.2-0.9 was observed in HBMEC total nuclei count for both proliferative and non-proliferative conditions. The fold difference in drug concentration required to achieve the specified Fa is indicated above each pair of columns.Verreault et al. BMC Cancer 2011, 11:124 14 of 18depending on the site of tumor growth (subcutaneous vsorthotopic). In the subcutaneous model, the liposomaltreatments increased the amount of Hoechst 33342staining in the tumor tissue (Figure 1c), while in theorthotopic tumors Hoechst 33342 staining was reduced(Figure 2c). As noted above, treatment effects were simi-lar if blood vessel morphology parameters were used asa measured endpoint. While initially surprising, theHoechst 33342 uptake data may actually be consistentwith restoration of the blood-brain barrier, which ismore impermeable to Hoechst 33342. It is well estab-lished that Hoechst 33342 is a p-glycoprotein substrate[42]. It does not accumulate in normal brain tissuebecause it cannot cross the blood brain barrier, but it ispresent in untreated orthotopic brain tumors whichexhibit leakier blood vessel. This idea was further con-firmed by Ktrans measurements, which strongly sug-gested a vasculature normalization induced byIrinophore C™. This interpretation suggests thatHoechst 33342 is not an appropriate marker for tumorperfusion in orthotopic glioma models, as it was pre-viously used in a s.c. tumor model [8]. It does, however,function as a permeability marker for perfused tumorassociated blood vessels, which is reduced upon normal-ization. The impact of vascular normalization on tumorperfusion in orthotopic GBM tumors could not beassessed in the present study because MRI Ktrans dataand Hoechst 33342 staining data are not direct mea-sures of perfusion in the brain tumor. However, dataobtained in the subcutaneous model suggest that treat-ment with liposomal drugs does not reduce tumor per-fusion, as measured by CD31/Hoechst 33342 doublestaining, and may even increase it, as suggested by dataobtained from Caelyx®-treated s.c. tumors. Studies tomeasure the delivery of a second drug that can cross theBBB in liposomal drug-treated tumors are underway andwill provide an indication of the impact of vascular nor-malization on vessel perfusion in the orthotopic model.The idea that liposomal formulations of anti-cancerdrugs, in addition to having a direct cytotoxic effect onthe tumor cells, may also act as through anti-angiogenicmechanisms is intriguing. It seems reasonable to suggestthat the extended drug release characteristics associatedwith the liposomal drug formulations used in this study[7,11] may have effects on blood vessels in a mannersimilar to metronomic dosing schedules - i.e. frequent,low dose administration of drugs with no prolongeddrug-free breaks [43]. Metronomic dosing is nowacknowledged to act specifically on the proliferatingendothelial cells of tumor blood vessels [44] and wasmore recently shown to improve tumor perfusion andto decrease hypoxia in a pancreatic tumor model [45].To examine this hypothesis, an in vitro assay was usedto evaluate the activity of irinotecan, doxorubicin andvincristine (the drugs encapsulated by liposome exam-ined in this study) against proliferating endothelial cells.The assay was adapted from one developed by Bocciet al. to examine the effects of metronomic drug expo-sure against endothelial cells [33]. Previous reports sug-gest that docetaxel and paclitaxel have potent activityagainst endothelial cells in an in vitro metronomic dos-ing regime [32,33,46], so these drugs were included inthe assay as positive controls. The effects of SN-38 werealso evaluated in the assay because SN-38 is a moreactive metabolite of irinotecan generated by tissue andplasma carboxylesterases in vivo [47,48]. Further it hasalready been established that following treatment withIrinophore C™, high levels of SN-38 are maintained inthe plasma compartment for extended time periods [7].SN-38 levels may play an important role in the anti-can-cer activity of Irinophore C™.The in vitro metronomic dosing assay presented inFigures 7 and 8 suggest that vincristine and SN-38, likethe taxanes (docetaxel or paclitaxel), are highly activeagainst proliferating endothelial cells (Figure 6a-b). Incontrast, free irinotecan has little specificity for prolifer-ating endothelial cells over non-proliferating cellsin vitro. The data for free vincristine corroborate theeffects on tumor vasculature seen with the liposomalform of the drug used here, while the results obtainedwith free irinotecan, which is not specific for proliferat-ing endothelial cells, is actually contradictory. IrinophoreC™ was the most active of the three liposomal formula-tions used. The results in Figure 7 and 8 would stronglysuggest that the activity of Irinophore C™may beexplained by the high plasma levels of SN-38 generatedfollowing administration of the formulation [7,16]. Thusit can be concluded from the studies presented herethat the active metabolite of irinotecan, SN-38, may bethe agent promoting vascular normalization in the mod-els used here.Interestingly, the in vitro assay suggests that doxorubi-cin should have little specificity on proliferatingendothelial cells, yet i.v. administration of Caelyx®resulted in effects on the tumor vasculature that werecomparable to those seen following administration ofIrinophore C™. The reasons for this are unclear at pre-sent but may be related to disruptions in the productionof hypoxia-induced VEGF caused by doxorubicin [49].Previous studies completed using the rat intracranial 9Ltumor model treated with a formulation of doxorubicincomparable to that used here [15] showed the presenceof vascular breakdown and hemorrhage 48 hours aftertreatment. In contrast, the results summarized herewere obtained using tumors harvested one week afterthe final treatment; thus the data here may reflect lateeffects on tumor vasculature. Further, 9L is a gliosar-coma cell line which exhibits a slower doubling timeVerreault et al. BMC Cancer 2011, 11:124 15 of 18(34.9 hrs [50]) than the U251MG glioblastoma cell line(20.9 hrs; data not shown) used in this study. Theresulting 9L tumors are also histological distinct [50]when compared to the U251MG model. These differ-ences will likely impact how tumors respond to agentscapable of promoting vascular normalization. Studiesassessing how vascular functions change in relationshipto tumor growth rate are currently being completed.ConclusionIn aggregate, data from this study indicates that liposomalformulations of irinotecan, doxorubicin and vincristineexert anti-angiogenic effects, as measured by endpointsassessing increases in mature blood vessels and improvedvascular function. The normalization of tumor vesselsappears to be transient in nature [36] but may create awindow where blood flow is improved, leading to anopportunity to improve drug delivery for other drugs. Thefact that all three formulations were therapeutically activein the orthotopic model suggests that vascular normaliza-tion did not prevent the drugs from accessing tumor cells,despite the fact that our interpretation of data obtainedfrom Hoechst 33342 suggests a reduction in vessel perme-ability. Data from our laboratory showed that once irinote-can is released from the lipid carrier, the drug and itsactive metabolite SN-38 are capable of crossing a normalblood-brain barrier (Verreault M, Strutt D, Masin D, Ana-ntha M, Waterhouse D, Yapp DT and Bally MB: Irino-phore C™, a lipid-based nanoparticulate formulation ofirinotecan, is more effective than free irinotecan whenused to treat an orthotopic glioblastoma model, submittedfor publication in March 2011). Vincristine was alsoshown to be able to cross a normal blood-brain barrier[51]. Thus, it can be speculated that vascular normaliza-tion would increase the delivery of drug that have disso-ciated from the liposome across the tumor vasculature,allowing higher levels of drug to diffuse into a greatervolume of tumor tissue. Studies assessing the conse-quences of liposomal drug-induced vascular normalizationon the delivery of a second drug capable of crossing theblood-brain barrier will provide important informationregarding the impact of tumor vessel permeability on drugdelivery. In the case of GBM, an obvious choice of such adrug is temozolomide. Pre-clinical studies to assess theimpact of Irinophore C™ treatments on the delivery oftemozolomide are currently on-going.Abbreviationsd-HMVEC: adult dermal human microvascular endothelial cells; Fa: fractionaffected; FBS: fetal bovine serum; GBM: glioblastoma; H&E: Hematoxylin andEosin; HBMEC: Human brain microvascular endothelial cells; i.v.: intravenous;s.c.: subcutaneous;AcknowledgementsThe research described in this original paper was supported by grantfunding from the Canadian Institutes of Health Research, the CancerResearch Society, Inc. and the National Cancer Institute of Canada (now theCanadian Cancer Society Research Institute). DTY was supported by RethinkBreast Cancer. The authors would like to acknowledge the staff andmanagement personnel of the Animal Research Center at the BC CancerAgency.Author details1Experimental Therapeutics, British Columbia Cancer Agency, 675 West10thAvenue, Vancouver, BC V5Z 1L3, Canada. 2Faculty of PharmaceuticalSciences, University of British Columbia, 2146 East Mall, Vancouver, BC V6T1Z3, Canada. 3Department of Pathology and Laboratory Medicine, Universityof British Columbia, 2211 Wesbrook Mall, Vancouver, BC V6T 2B5, Canada.4Center for Drug Research and Development, Vancouver, BC V6T 1Z4,Canada. 5UBC MRI Research Center, 2221 Wesbrook Mall, Vancouver, BC V6T2B5, Canada.Authors’ contributionsMV carried out all parts of the experimental manipulations, data analysis anddraft of manuscript. DS and DM were involved in the implantation of s.c.and orthotopic tumors and monitoring of the animals. MA and DW wereinvolved in the development of Irinophore C™ formulation. AY and PK werepart of MRI-DCE data acquisition and analysis. MBB and DTY were involvedin the conception of the study, participated in its design and helped to draftthe manuscript. All authors read and approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Received: 5 October 2010 Accepted: 8 April 2011 Published: 8 April 2011References1. Blasberg RG, Kobayashi T, Horowitz M, Rice JM, Groothuis D, Molnar P,Fenstermacher JD: Regional blood flow in ethylnitrosourea-induced braintumors. Ann Neurol 1983, 14:189-201.2. Groothuis DR, Pasternak JF, Fischer JM, Blasberg RG, Bigner DD, Vick NA:Regional measurements of blood flow in experimental RG-2 rat gliomas.Cancer Res 1983, 43:3362-7.3. Vajkoczy P, Schilling L, Ullrich A, Schmiedek P, Menger MD:Characterization of angiogenesis and microcirculation of high-gradeglioma: an intravital multifluorescence microscopic approach in theathymic nude mouse. J Cereb Blood Flow Metab 1998, 18:510-20.4. Baish JW, Gazit Y, Berk DA, Nozue M, Baxter LT, Jain RK: Role of tumorvascular architecture in nutrient and drug delivery: an invasionpercolation-based network model. Microvasc Res 1996, 51:327-46.5. Yuan F, Salehi HA, Boucher Y, Vasthare US, Tuma RF, Jain RK: Vascularpermeability and microcirculation of gliomas and mammary carcinomastransplanted in rat and mouse cranial windows. Cancer Res 1994,54:4564-8.6. Vajkoczy P, Menger MD: Vascular microenvironment in gliomas.J Neurooncol 2000, 50:99-108.7. Ramsay EC, Anantha M, Zastre J, Meijs M, Zonderhuis J, Strutt D, Webb MS,Waterhouse D, Bally MB: Irinophore C: a liposome formulation ofirinotecan with substantially improved therapeutic efficacy against apanel of human xenograft tumors. Clin Cancer Res 2008, 14:1208-17.8. Baker JH, Lam J, Kyle AH, Sy J, Oliver T, Co SJ, Dragowska WH, Ramsay E,Anantha M, Ruth TJ, Adam MJ, Yung A, Kozlowski P, Minchinton AI, Ng SS,Bally MB, Yapp DT: Irinophore C, a novel nanoformulation of irinotecan,alters tumor vascular function and enhances the distribution of 5-fluorouracil and doxorubicin. Clin Cancer Res 2008, 14:7260-71.9. Baffert F, Le T, Sennino B, Thurston G, Kuo CJ, Hu-Lowe D, McDonald DM:Cellular changes in normal blood capillaries undergoing regression afterinhibition of VEGF signaling. Am J Physiol Heart Circ Physiol 2006, 290:H547-59.10. Jain RK: Normalizing tumor vasculature with anti-angiogenic therapy: anew paradigm for combination therapy. Nat Med 2001, 7:987-9.Verreault et al. BMC Cancer 2011, 11:124 16 of 1811. Mayer LD, Masin D, Nayar R, Boman NL, Bally MB: Pharmacology ofliposomal vincristine in mice bearing L1210 ascitic and B16/BL6 solidtumours. Br J Cancer 1995, 71:482-8.12. Mabeta P, Pepper MS: A comparative study on the anti-angiogeniceffects of DNA-damaging and cytoskeletal-disrupting agents.Angiogenesis 2009, 12:81-90.13. Chen Q, Tong S, Dewhirst MW, Yuan F: Targeting tumor microvesselsusing doxorubicin encapsulated in a novel thermosensitive liposome.Mol Cancer Ther 2004, 3:1311-7.14. Arnold RD, Mager DE, Slack JE, Straubinger RM: Effect of repetitiveadministration of Doxorubicin-containing liposomes on plasmapharmacokinetics and drug biodistribution in a rat brain tumor model.Clin Cancer Res 2005, 11:8856-65.15. Zhou R, Mazurchuk R, Straubinger RM: Antivasculature effects ofdoxorubicin-containing liposomes in an intracranial rat brain tumormodel. Cancer Res 2002, 62:2561-6.16. Ramsay E, Alnajim J, Anantha M, Zastre J, Yan H, Webb M, Waterhouse D,Bally M: A novel liposomal irinotecan formulation with significant anti-tumour activity: use of the divalent cation ionophore A23187 andcopper-containing liposomes to improve drug retention. Eur J PharmBiopharm 2008, 68:607-17.17. Mayer LD, Bally MB, Loughrey H, Masin D, Cullis PR: Liposomal vincristinepreparations which exhibit decreased drug toxicity and increasedactivity against murine L1210 and P388 tumors. Cancer Res 1990,50:575-9.18. Tomayko MM, Reynolds CP: Determination of subcutaneous tumor size inathymic (nude) mice. Cancer Chemother Pharmacol 1989, 24:148-54.19. Lyng H, Dahle GA, Kaalhus O, Skretting A, Rofstad EK: Measurement ofperfusion rate in human melanoma xenografts by contrast-enhancedmagnetic resonance imaging. Magn Reson Med 1998, 40:89-98.20. Tofts PS, Brix G, Buckley DL, Evelhoch JL, Henderson E, Knopp MV,Larsson HB, Lee TY, Mayr NA, Parker GJ, Port RE, Taylor J, Weisskoff RM:Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: standardized quantities andsymbols. J Magn Reson Imaging 1999, 10:223-32.21. Low J, Huang S, Blosser W, Dowless M, Burch J, Neubauer B, Stancato L:High-content imaging characterization of cell cycle therapeutics throughin vitro and in vivo subpopulation analysis. Mol Cancer Ther 2008,7:2455-63.22. Stravopodis DJ, Karkoulis PK, Konstantakou EG, Melachroinou S,Lampidonis AD, Anastasiou D, Kachrilas S, Messini-Nikolaki N, Papassideri IS,Aravantinos G, Margaritis LH, Voutsinas GE: Grade-dependent effects oncell cycle progression and apoptosis in response to doxorubicin inhuman bladder cancer cell lines. Int J Oncol 2009, 34:137-60.23. Foroodi F, Duivenvoorden WC, Singh G: Interactions of doxycycline withchemotherapeutic agents in human breast adenocarcinoma MDA-MB-231 cells. Anticancer Drugs 2009, 20:115-22.24. Saleh EM, El-Awady RA, Abdel Alim MA, Abdel Wahab AH: Alteredexpression of proliferation-inducing and proliferation-inhibiting genesmight contribute to acquired doxorubicin resistance in breast cancercells. Cell Biochem Biophys 2009, 55:95-105.25. Landberg G, Tan EM, Roos G: Flow cytometric multiparameter analysis ofproliferating cell nuclear antigen/cyclin and Ki-67 antigen: a new view ofthe cell cycle. Exp Cell Res 1990, 187:111-8.26. Bernsen HJ, Rijken PF, Hagemeier NE, van der Kogel AJ: A quantitativeanalysis of vascularization and perfusion of human glioma xenografts atdifferent implantation sites. Microvasc Res 1999, 57:244-57.27. Bernsen HJ, Rijken PF, Peters H, Raleigh JA, Jeuken JW, Wesseling P, van derKogel AJ: Hypoxia in a human intracerebral glioma model. J Neurosurg2000, 93:449-54.28. Kamoun WS, Ley CD, Farrar CT, Duyverman AM, Lahdenranta J, Lacorre DA,Batchelor TT, di Tomaso E, Duda DG, Munn LL, Fukumura D, Sorensen AG,Jain RK: Edema control by cediranib, a vascular endothelial growth factorreceptor-targeted kinase inhibitor, prolongs survival despite persistentbrain tumor growth in mice. J Clin Oncol 2009, 27:2542-52.29. Winkler F, Kozin SV, Tong RT, Chae SS, Booth MF, Garkavtsev I, Xu L,Hicklin DJ, Fukumura D, di Tomaso E, Munn LL, Jain RK: Kinetics of vascularnormalization by VEGFR2 blockade governs brain tumor response toradiation: role of oxygenation, angiopoietin-1, and matrixmetalloproteinases. Cancer Cell 2004, 6:553-63.30. Baluk P, Hashizume H, McDonald DM: Cellular abnormalities of bloodvessels as targets in cancer. Curr Opin Genet Dev 2005, 15:102-11.31. Yang Z, Suzuki R, Daniels SB, Brunquell CB, Sala CJ, Nishiyama A: NG2 glialcells provide a favorable substrate for growing axons. J Neurosci 2006,26:3829-39.32. Bocci G, Falcone A, Fioravanti A, Orlandi P, Di Paolo A, Fanelli G, Viacava P,Naccarato AG, Kerbel RS, Danesi R, Del Tacca M, Allegrini G: Antiangiogenicand anticolorectal cancer effects of metronomic irinotecanchemotherapy alone and in combination with semaxinib. Br J Cancer2008, 98:1619-29.33. Bocci G, Nicolaou KC, Kerbel RS: Protracted low-dose effects on humanendothelial cell proliferation and survival in vitro reveal a selectiveantiangiogenic window for various chemotherapeutic drugs. Cancer Res2002, 62:6938-43.34. Maeda H, Bharate GY, Daruwalla J: Polymeric drugs for efficient tumor-targeted drug delivery based on EPR-effect. Eur J Pharm Biopharm 2009,71:409-19.35. O’Connor JP, Jackson A, Parker GJ, Jayson GC: DCE-MRI biomarkers in theclinical evaluation of antiangiogenic and vascular disrupting agents. Br JCancer 2007, 96:189-95.36. Jain RK: Normalization of tumor vasculature: an emerging concept inantiangiogenic therapy. Science 2005, 307:58-62.37. Le Serve AW, Hellmann K: Vascular changes in tumours after treatmentwith ICRF 159. Br J Pharmacol 1971, 43:457P-458P.38. Vajkoczy P, Menger MD: Vascular microenvironment in gliomas. CancerTreat Res 2004, 117:249-62.39. Deane BR, Lantos PL: The vasculature of experimental brain tumours. Part1. A sequential light and electron microscope study of angiogenesis.J Neurol Sci 1981, 49:55-66.40. Sorensen AG, Batchelor TT, Zhang WT, Chen PJ, Yeo P, Wang M,Jennings D, Wen PY, Lahdenranta J, Ancukiewicz M, di Tomaso E, Duda DG,Jain RK: A “vascular normalization index” as potential mechanisticbiomarker to predict survival after a single dose of cediranib inrecurrent glioblastoma patients. Cancer Res 2009, 69:5296-300.41. Ma J, Li S, Reed K, Guo P, Gallo JM: Pharmacodynamic-mediated effects ofthe angiogenesis inhibitor SU5416 on the tumor disposition oftemozolomide in subcutaneous and intracerebral glioma xenograftmodels. J Pharmacol Exp Ther 2003, 305:833-9.42. Shapiro AB, Corder AB, Ling V: P-glycoprotein-mediated Hoechst 33342transport out of the lipid bilayer. Eur J Biochem 1997, 250:115-21.43. Hanahan D, Bergers G, Bergsland E: Less is more, regularly: metronomicdosing of cytotoxic drugs can target tumor angiogenesis in mice. J ClinInvest 2000, 105:1045-7.44. Browder T, Butterfield CE, Kraling BM, Shi B, Marshall B, O’Reilly MS,Folkman J: Antiangiogenic scheduling of chemotherapy improvesefficacy against experimental drug-resistant cancer. Cancer Res 2000,60:1878-86.45. Cham KK, Baker JH, Takhar KS, Flexman JA, Wong MQ, Owen DA, Yung A,Kozlowski P, Reinsberg SA, Chu EM, Chang CW, Buczkowski AK,Chung SW, Scudamore CH, Minchinton AI, Yapp DT, Ng SS: Metronomicgemcitabine suppresses tumour growth, improves perfusion, andreduces hypoxia in human pancreatic ductal adenocarcinoma. Br JCancer 2010, 103:52-60.46. Vacca A, Ribatti D, Iurlaro M, Merchionne F, Nico B, Ria R, Dammacco F:Docetaxel versus paclitaxel for antiangiogenesis. J Hematother Stem CellRes 2002, 11:103-18.47. Slatter JG, Su P, Sams JP, Schaaf LJ, Wienkers LC: Bioactivation of theanticancer agent CPT-11 to SN-38 by human hepatic microsomalcarboxylesterases and the in vitro assessment of potential druginteractions. Drug Metab Dispos 1997, 25:1157-64.48. Lavelle F, Bissery MC, Andre S, Roquet F, Riou JF: Preclinical evaluation ofCPT-11 and its active metabolite SN-38. Semin Oncol 1996, 23:11-20.49. Duyndam MC, van Berkel MP, Dorsman JC, Rockx DA, Pinedo HM, Boven E:Cisplatin and doxorubicin repress Vascular Endothelial Growth Factorexpression and differentially down-regulate Hypoxia-inducible Factor Iactivity in human ovarian cancer cells. Biochem Pharmacol 2007,74:191-201.50. Asai A, Shibui S, Barker M, Vanderlaan M, Gray JW, Hoshino T: Cell kineticsof rat 9L brain tumors determined by double labeling with iodo- andbromodeoxyuridine. J Neurosurg 1990, 73:254-8.Verreault et al. BMC Cancer 2011, 11:124 17 of 1851. Wang F, Zhou F, Kruh GD, Gallo JM: Influence of blood-brain barrier effluxpumps on the distribution of vincristine in brain and brain tumors.Neuro Oncol 2010, 12:1043-9.Pre-publication historyThe pre-publication history for this paper can be accessed here: this article as: Verreault et al.: Vascular normalization in orthotopicglioblastoma following intravenous treatment with lipid-basednanoparticulate formulations of irinotecan (Irinophore C™™),doxorubicin (Caelyx®®) or vincristine. BMC Cancer 2011 11:124.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at et al. BMC Cancer 2011, 11:124 18 of 18


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items