Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evaluating the extravascular penetration and distribution of anti-cancer drugs using a novel application… Kyle, Alastair Hugh 2007

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

Item Metadata

Download

Media
831-ubc_2007-318065.pdf [ 32.11MB ]
Metadata
JSON: 831-1.0100652.json
JSON-LD: 831-1.0100652-ld.json
RDF/XML (Pretty): 831-1.0100652-rdf.xml
RDF/JSON: 831-1.0100652-rdf.json
Turtle: 831-1.0100652-turtle.txt
N-Triples: 831-1.0100652-rdf-ntriples.txt
Original Record: 831-1.0100652-source.json
Full Text
831-1.0100652-fulltext.txt
Citation
831-1.0100652.ris

Full Text

E V A L U A T I N G T H E EXTRA VASCULAR PENETRATION A N D DISTRIBUTION OF ANTI-CANCER DRUGS USING A NOVEL APPLICATION OF T H E MULTILAYERED CELL CULTURE MODEL by ALASTAIR H U G H KYLE B.Sc, M c G i l l University, 1993 M.Sc , University of British Columbia, 1999 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF G R A D U A T E STUDIES (Experimental Medicine) THE UNIVERSITY OF BRITISH COLUMBIA JULY 2007 © Alastair Hugh Kyle, 2007 ii ABSTRACT The inefficient distribution of anti-cancer drugs in the tumour extravascular compartment may be an important mechanism of drug resistance in solid tumours. However, gaining an understand-ing of diffusion limitations of existing drugs and developing new drugs with improved tumour penetration is limited by a lack of techniques with which to evaluate drugs. The purpose of this research project was to assess the extravascular penetration and distribution of anti-cancer drugs using a novel application of the multilayered cell culture (MCC) model. MCC is a planar ana-logue of spheroidal cell culture, in which tumour cells are instead grown into discs of tissue. Due to their 3-D conformation they model many characteristics of the tumour extravascular compart-ment. A unique property of MCC is that it possesses two populations of rapidly proliferating cells, one on each side of the culture, that are separated by a known thickness of tissue. In tumours and spheroids proliferation status falls off with distance from the vasculature (tumours) or tissue edge (spheroids), which in turn modifies cellular response to drugs. Hence, visualizing the distribution of a drug's effect within these tissues cannot be used to directly determine its actual distribution. The hypothesis of this project was that the symmetrical proliferation seen in MCCs could be applied to measure the ability of anti-cancer drugs to penetrate and distribute within tissue using a biological endpoint. Based on this hypothesis an effect-based assay was devised and applied to assess the tissue distribution of a selection of anti-cancer drugs from several distinct classes. Using the assay, drug exposure to cells 150 um into tissue (a diffusion distance commonly observed in solid tumours) was estimated to vary from as low as 3-10% of the reservoir exposure (anthracyclines) to cases where no reduction in exposure could be detected (5-FU and cisplatin). The MCC-based assay was validated by comparing the predicted drug distri-butions with direct visualization of the drugs themselves. Results from this study suggest that an MCC effect-based assay could be used as a simple drug penetration screen in the development of new anti-cancer drugs. T A B L E OF CONTENTS A B S T R A C T i i T A B L E OF CONTENTS i i i LIST OF TABLES vi i LIST OF FIGURES vi i i LIST OF ABBREVIATIONS & DEFINITIONS xi 1. INTRODUCTION 1 1.1 T U M O U R M I C R O E N V I R O N M E N T 2 ARCHITECTURE 3 FACTORS THAT DEFINE THE TUMOUR MICROENVIRONMENT ' 5 Diffusion limitations 5 Variable vessel perfusion 6 Harsh conditions & aggressive phenotypes 6 M O D U L A T I O N OF THERAPY 7 1.2 D R U G P E N E T R A T I O N 8 FACTORS INFLUENCING TUMOUR DRUG DISTRIBUTION: SUPPLY A N D D E M A N D 8 T U M O U R MAPPING 10 Direct imaging of drugs and their adducts 10 Mapping drug effect in tumours 11 Limitations of tumour mapping for evaluation of drug penetration 11 1.3 In v i t r o T U M O U R M O D E L S 1 2 SPHEROIDAL CELL CULTURE 12 Modelling solid tumours 12 Effect of tumour microenvironment on response 12 Drug penetration 13 MULTILAYERED CELL CULTURE 13 MCC-based flux assay .13 Simulating drug distribution 16 Visualizing drug fluorescence in MCC 16 1.4 P R O J E C T O U T L I N E : D E V E L O P I N G A N O V E L M C C - B A S E D A S S A Y F O R A S S E S S I N G D R U G D I S T R I B U T I O N U S I N G A B I O L O G I C A L E N D P O I N T 18 EFFECT-BASED DRUG DISTRIBUTION ASSAY 18 Choice of boundary conditions 18 Procedure 19 Interpreting results 20 iv PROJECT HYPOTHESIS 21 DEVELOPMENT AND VALIDATION OF THE M C C EFFECT-BASED ASSAY 21 1.5 OUTLINE OF PRESENTED WORK 22 SUITABILITY OF THE S-PHASE MARKER BRDURD AS BIOLOGICAL END-POINT 22 Plan of work 22 ANTHRACYCLINE DISTRIBUTION IN M C C : EFFECT-BASED ASSAY VERSUS DIRECT VISUALIZATION ..23 Plan of work 23 TAXANE DISTRIBUTION IN M C C & TUMOURS: EFFECT-BASED ASSAY & DIRECT VISUALIZATION 24 Plan of work 25 ANTI-CANCER DRUG SURVEY: EFFECT-BASED ASSAY 26 Vinca alkaloids 26 Antimetabolites 26 Others 27 Plan of work 27 2. M A T E R I A L S & M E T H O D S 28 2.1 CELL CULTURE 28 MONOLAYER CULTURE 28 MULTILAYERED CELL CULTURE 28 2.2 B R D U R D - P E N E T R A T I O N IN M C C A N D XENOGRAFTS 29 MICE AND TUMORS 29 TUMOUR PENETRATION ASSAY 29 MCC-BASED BRDURD PENETRATION ASSAY 29 PHARMACOKINETIC ASSAY 30 HIGH PRESSURE LIQUID CHROMATOGRAPHY 30 BRDURD IMMUNOHISTOCHEMISTRY 31 IMAGE ACQUISITION 31 IMAGE ANALYSIS: TUMOURS 31 IMAGE ANALYSIS: M C C s 32 2.3 EFFECT-BASED M C C D R U G DISTRIBUTION ASSAY 32 M C C CONTROL ASSAY 33 BRDURD IMMUNOHISTOCHEMISTRY 33 IMAGE ACQUismoN b 34 IMAGE ACQUISITION^ 34 IMAGE ANALYSIS 34 2.4 A N T H R A C Y C L I N E FLUX EXPERIMENTS 34 FLUX ASSAY34 V HIGH PRESSURE LIQUID CHROMATOGRAPHY 35 Sample Preparation 35 Analysis 35 2.5 3 H-TAXANE AND TAXANE-TUMOUR XENOGRAFT WORK 35 REAGENTS 35 MICE AND TUMOURS 36 RADIOLABELLED TAXANES IN M C C S 37 IMMUNOHISTOCHEMISTRY 38 IMAGE ANALYSIS 38 3. BROMODEOXYURIDINE 39 3.1 SUMMARY 39 3.2 RESULTS 39 BRDURD PENETRATION INTO TUMOR XENOGRAFTS 39 BRDURD PHARMACOKINETICS IN SCID MICE 41 BRDURD PENETRATION INTO MCC 42 3.3 DISCUSSION 45 4. ANTHRACYCLINES 47 4.1 SUMMARY 47 4.2 RESULTS 47 PRELIMINARY FINDINGS: FLUX-BASED EVALUATION OF ANTHRACYCLINE DIFFUSION THROUGH MCCS 47 EFFECT BASED EVALUATION OF ANTHRACYCLINE PENETRATION IN MCCs 50 FLUORESCENCE BASED EVALUATION OF ANTHRACYCLINE DISTRIBUTION IN MCCs 56 INHIBITION OF ACIDIC SEQUESTRATION 57 4.3 DISCUSSION 58 5. T A X A N E S , 61 5.1 SUMMARY 61 5.2 RESULTS 61 EFFECT-BASED EVALUATION OF TAXANE TISSUE DISTRIBUTION IN MCCs 61 3 H-TAXANE DISTRIBUTION IN M C C S 64 3 H-TAXANE DISTRIBUTION IN TUMOUR XENOGRAFTS 65 COMPARISON OF XENOGRAFT AND MCC-BASED 3 H-TAXANE DATA 67 EFFECT OF TAXANES ON S-PHASE CELL DISTRIBUTION IN TUMOUR XENOGRAFTS 69 5.3 DISCUSSION 70 vi 6. SURVEY OF THE PENETRATION AND DISTRIBUTION OF SOME COMMON ANTI-CANCER DRUGS 73 6.1 SUMMARY 73 6.2 RESULTS .73 V l N C A ALKALOIDS 73 ANTIMETABOLITES 76 PLATINUM AGENTS - CISPLATIN 79 B-LEOMYCIN81 ETOPOSIDE 83 MITOMYCIN C 85 ALKYLATING AGENTS 87 HYPOXIC PROFILE 88 6.3 DISCUSSION 90 7. SUMMARY AND FUTURE DIRECTIONS 91 REFERENCES: 96 v i i LIST OF TABLES TABLE 1-1 Diffusion/perfusion limited gradients in solid tumours and their effect on therapy 7 TABLE 1-2 Factors which wi l l influence a drug's ability to penetrate and distribute within solid tumours 9 TABLE 4-1. Drug exposure to the near side relative to the far side of M C C s as estimated from the effect-based assay or drug derived fluorescence 57 TABLE 7-1 Summary of results obtained from the M C C effect-based assay for evaluation of drug distribution 92 TABLE 7-2 Summary of D N A and microtubule interactions and cell cycle effects of the drugs tested using the effect-based assay 93 TABLE 7-3 Summary of human plasma pharmacokinetic and physicochemical properties of the drugs tested using the effect-based assay 94 LIST OF FIGURES FIGURE 1-1 Examples of two tumour architectures commonly observed in experimental tumours grown subcutaneously in mice 4 FIGURE 1-2 Example of a clinical human breast cancer 5 FIGURE 1-3 A | Examples of two possible routes for a hypothetical anthracycline-type drug that accumulates through DNA intercalation. B | Schematic of known cellular interactions of the anthracycline doxorubicin 10 FIGURE 1-4 Multilayered cell culture (MCC) growth apparatus and histology 13 FIGURE 1-5 Schematics of two types of apparatus used for MCC-based flux experiments and typical data produced by the assay ; 14 FIGURE 1-6 Schematic of the process of drug build-up within an MCC during flux experiments using the dual-reservoir configuration 15 FIGURE 1-7 Apparatus used to exposed MCC to drug from just one side 16 FIGURE 1-8 Drug derived fluorescent profiles observed in MCC following 1-hour exposure using the closed-off apparatus 17 FIGURE 1-9 Symmetrical S-phase cell distribution seen in MCC as detected by immunohistochemical staining of BrdUrd incorporated into DNA 18 FIGURE 1-10 Examples of two possible boundary situations: closed-off versus open 19 FIGURE 1-11 Outline of the assay proposed to evaluate drug distributions using the symmetrical proliferation observed in MCC 20 FIGURE 1-12 Estimating drug distribution based on comparison of dose response curves on exposed versus far sides of MCCs 21 FIGURE 1-13 Chemical structures of the anthracyclines to be evaluated using the effect-based assay 24 FIGURE 2-1 Schematic of growth vessel and drug penetration apparatus used for MCC-based work 28 FIGURE 2-2 Workflow for BrdUrd tumour labelling experiments designed to examine the distribution of labelled S-phase cells in relation to vasculature 29 FIGURE 2-3 Workflow for BrdUrd MCC labelling experiment designed to examine the distribution of labelled S-phase cells in relation to culture edge 30 FIGURE 2-4 Outline of the proposed effect-based assay for evaluation of drug distribution 32 FIGURE 2-5 Radiolabelled taxane tumour mapping protocol 36 FIGURE 2-6 Protocol for mapping the effect of taxanes on S-phase cell distribution tumours 37 FIGURE 2-7 Radiolabelled taxane MCC mapping protocol 37 FIGURE 3-1 SiHa tumour cryosections showing S-phase labelled cells (greyscale) relative to vasculature (magenta) as a function of BrdUrd dose 40 FIGURE 3-2 BrdUrd plasma kinetics in SCID mice after intraperitonal administration of 100 mg/kg 41 FIGURE 3-3 Greyscale images of 3-day-old MCC showing the distribution of S-phase labelled cells as a function of BrdUrd concentration 42 FIGURE 3-4 Greyscale images of 4-day-old MCC showing distribution of BrdUrd labelled S-phase cells after extended exposure from both sides under various conditions 44 ix FIGURE 4-1 Flux data showing the rate of appearance of anthracyclines and their metabolites in the receiving reservoir following diffusion through SiHa M C C s 49 FIGURE 4-2 Anthracycline fluorescence in 10 um cryosections taken from SiHa M C C s exposed to 100 u M drug for 3 hours under conditions used for flux experiments 50 FIGURE 4-3 Proof of principle of BrdUrd-based drug penetration assay 51 FIGURE 4-4 Comparison of BrdUrd labelling on far-side relative near side of M C C s following 45 min and 4 h closures 52 FIGURE 4-5. Distribution of S-phase cells in M C C s 1 day after a 1-h exposure to increasing concentrations of EPI (0.3 to 10 uM) from one side 53 FIGURE 4-6. Comparison of the effect of the four anthracyclines on the first 30 um of tissue on the near side and far side of M C C s 1 day after a 1-h exposure 54 FIGURE 4-7. Distribution of S-phase cells in M C C s 3 days after a 1-h exposure from one side to the four anthracyclines at their peak human plasma levels 55 FIGURE 4-8. Anthracycline fluorescence in M C C s following a 1-h exposure at 10 u M drug from one side 56 FIGURE 4-9 Effect of inhibition of acid sequestration via ammonium chloride on the tissue distribution of drug derived fluorescence in HCT-116 M C C s 58 FIGURE 4-10 Effect of inhibition of acidic sequestration on BrdUrd labelling in HCT-116 M C C after exposure to 1 u M daunorubicin for 1-hour from one side 58 FIGURE 5-1 S-phase cells in HCT-116 M C C s 2 days after a 1-hour drug exposure from one side. .62 FIGURE 5-2 Analysis of drug effect on the rapidly proliferating cells of the first 30 um of tissue on the near and far side of M C C s for paclitaxel and docetaxel 63 FIGURE 5-3 Image of 3H-paclitaxel distribution in HCT-116 M C C following a 1-hour exposure to 3 uM drug from one side 64 FIGURE 5-4 3H-taxane distributions in HCT-116 M C C s following 0.3 & 3 u M exposures for 1 & 2 hours to paclitaxel and docetaxel 65 FIGURE 5-5 Composite image of 3H-paclitaxel distribution (black) and CD31 positive vasculature (red) in an HCT-116 tumour xenograft 2 hours post i.v. injection of 10 m g / k g 3H-paclitaxel 66 FIGURE 5-6 3H-taxane distribution relative nearest blood vessel in HCT-116 tumour xenografts. .67 FIGURE 5-7 Summary of 3H-taxane peak tissue levels, total radioactivity and depth into tissue where radioactivity falls to half-maximum 68 FIGURE 5-8 S-phase cells seen in HCT-116 tumour xenografts 72 hours following i.p. administration of paclitaxel and docetaxel at approximately their maximum tolerated doses 69 FIGURE 5-9 Analysis of S-phase cell distribution HCT-116 tumour xenografts following i.p. administration of paclitaxel and docetaxel 70 FIGURE 6-1 Vinca alkaloid data: S-phase cells 1 day after exposure to vincristine, vinblastine or vinorelbine for 1 hour to one side 74 FIGURE 6-2 Vinca alkaloids analysis 75 FIGURE 6-3 Gemcitabine data: S-phase cells 1 & 3 days after exposure to 3-90 u M gemcitabine for 1 hour from one side 76 FIGURE 6-4 Gemcitabine analysis 77 X FIGURE 6-5 5-FU & cytarabine data: S-phase cells 1 day after exposure to 5-FU or cytarabine for 1 hour to one side 78 FIGURE 6-6 Cisplatin data: S-phase cells 1 & 3 days after exposure to 10-300 uM cisplatin for 1 hour from one side 79 FIGURE 6-7 Cisplatin analysis 80 FIGURE 6-8 Bleomycin data: S-phase cells 1 day after exposure to bleomycin for 1 hour from one side 81 FIGURE 6-9 Bleomycin analysis 82 FIGURE 6-10 Etoposide data: S-phase cells 1 day after exposure to etoposide for 1 hour from one side 83 FIGURE 6-11 Etoposide analysis 84 FIGURE 6-12 Mitomycin Cdata: S-phase cells 1 & 3 days after exposure to mitomycin C for 1 hour from one side 85 FIGURE 6-13 Mitomycin C analysis 86 FIGURE 6-14 Alkylating agents data: S-phase cells 3 days after exposure to bcnu or thiotepa for 1 hour to one side 87 FIGURE 6-15 Hypoxic profile: Pimonidazole immunostaining profile in 150 um thick HCT-116 MCC 88 FIGURE 6-16 Hypoxic profile analysis 89 xi LIST OF ABBREVIATIONS & DEFINITIONS 5-FU 5-fluorouracil AUC area under the curve BGS™ bovine growth serum (FBS replacement) BrdUrd 5-bromo-2-deoxy-uridine Cremophor® El Polyoxyl 35 Castor Oil (vehicle used in paclitaxel formulation) DiOC7(3) a carbocyanine, fluorescent molecule used as blood perfusion marker DOX doxorubicin (adriamycin) DAU daunorubicin EPI epirubicin FBS foetal bovine serum 3 H tritium Hypoxia tissue having low a oxygen status IC50 drug concentration that produces a 50% decrease in proliferation MCC multilayered cell culture - cells grown as discs of tissue MIT mitoxantrone PTFE polytetrafluoroethylene (Teflon®) SD standard deviation SE standard error of the mean Spheroids multicellular spheroids - cells grown as spherical aggregates Xenografts tumour cells of non-mouse origin grown in mice Chapter 1 - Introduction 1 1. INTRODUCTION The role played by the tumour microenvironment in limiting the extravascular distribution of anti-cancer drugs has been largely ignored in the development of most new and existing agents (Minchinton and Tannock 2006). The most commonly measured parameter characterising drug kinetics within solid tumours is net tumour drug accumulation. However, critically, this parame-ter does not differentiate between drugs that distribute throughout the tissue versus those that are confined to the first few cells surrounding tumour vasculature. Through studies in solid tumours and multicellular spheroids it has become clear that the tumour microenvironment plays a critical role in modifying the efficacy of anti-cancer drugs (see sections 1.1 T U M O U R M I C R O E N V I R O N M E N T and 1.2 D R U G P E N E T R A T I O N ) . Unregulated proliferation and survival of tumour cells result in increased distances between tumour blood vessels producing gradients in drugs, oxygen, nutrients, proliferation and acidity away from blood vessels, all of which can modify the success of anti-cancer drugs that are generally selected from monolayer-centric studies. Multicellular spheroids are now often applied to integrate these factors and determine the activity of new anti-cancer drugs in an environment that is more representative of solid tumours than monolayer cell culture (see section 1.3 - IN VITRO T U M O U R M O D E L S ) . However, when a drug fails in spheroids it is often difficult to ascertain what role tissue penetration may have played in limiting its efficacy. The most commonly used methods of assessing the tissue penetration of a drug are either through visualisation of the drug itself within tissue cryosections (see section 1.2 - direct imaging of drugs) or through characterisation of drug flux through slabs of tissue followed by analysis and mathematical modelling to predict tissue pharmacokinetics (see section 1.3 - simulating drug distribution). However both these approaches have certain drawbacks. Imaging techniques (e.g. autoradiography, drug derived fluorescence or immunodetection) often can't differentiate between parent drug and inactive metabolites that may be smaller and diffuse more easily and drug derived fluorescence can be modified by cell interactions (e.g. binding, acidic sequestration). Immunodetection also requires the drug to be immobilised to avoid redistribution and washout during the immunostaining process. Additionally, direct imaging produces a discrete snapshot in time and to determine overall exposure to drug requires a sequence of snapshots spanning the entire process of drug build-up and subsequent washout. Flux-based studies coupled with mathematical modelling can in theory solve many of these issues. However, successful applica-tion of flux data to predict the tissue pharmacokinetics of a drug requires the full characterisation of all relevant drug-tissue interactions that may contribute in determining its tissue penetration. The goal of this project was to develop an assay that would be able to assess drug penetration using a biological endpoint. Potential benefits from such an approach would be that it could be fast and simple, not requiring the drug to be visualizable, and also be insensitive to inactive Chapter 1 - Introduction 2 metabolites. It might also be able to integrate drug exposure over the entire build-up and wash out process and could potentially produce meaningful results without requiring the full charac-terization of all relevant drug-cell interactions. Such an assay could thereby allow for the tissue penetration-based screening of new drugs for which physicochemical parameters had not been completely assessed. The purpose of this research was to investigate the use of a novel application of multilayered cell culture (MCC) in the assessment of the extravascular penetration and distribution of anti-cancer drugs using a biological endpoint. MCC is the planar analogue of spheroidal cell culture, in which tumour cells are grown into discs of tissue rather than spherical aggregates. Due to their growth in a 3-D conformation spheroids and MCCs can model many characteristics of the tumour extravascular compartment (see section 1.3 - IN VITRO TUMOUR MODELS). MCC was originally developed to allow measurement of the flux of drugs through tumour-like tissue (see section 1.3 - MULTILAYERED CELL CULTURE). This project investigates a novel application of MCC to directly assess a drug's ability to distribute within tissue that does not require detection of the drug itself by making use of a unique property of MCC, that it possesses two populations of rapidly proliferating cells, one on each edge of the culture, that are separated by a known thickness (see section 1.4). In tumours and spheroids, proliferation falls off with distance from the vasculature/depth into tissue, which in turn affects cellular response to drugs (see section 1.1 and 1.3 - SPHEROIDAL CELL CULTURE). Hence, visualizing the distribution of a drug's effect within these tissues cannot be used to directly determine its actual distribution. The hypothesis of this project was that the symmetrical proliferation seen in MCC could be used as a simple means of measur-ing the ability of a drug to penetrate and distribute in tissue. That is, exposing just one side of an MCC to a drug and then comparing the effect on the two proliferating populations will allow for evaluation of drug penetration without the need for direct visualization or mathematical model-ling of tissue pharmacokinetics based on MCC-derived flux-data. 1.1 TUMOUR MICROENVIRONMENT While cancer cells grown in a petri dish can be selectively targeted and killed, when grown as solid tumours achieving cure often becomes an intractable problem. One of the reasons for this is the unusual environment that is created by the combination of sustained expansion of the tumour extravascular compartment and subsequent imperfect neovascularization. Deregulated prolifera-tion and survival of tumour cells result in increased separation of blood vessels (Thomlinson and Gray 1955; Rubin and Casarett 1966) and unstable perfusion (Brown 1979; Chaplin et al. 1986; Minchinton et al. 1990; Durand and LePard 1995), in turn leading to a reduction in the distribu-tion and removal of molecules supplied by the blood 0ain 1987; Jain 1997). Tumour cells can be located up to 15-20 cells away from the nearest blood vessel (more in some cases) while in most normal tissue cells are within a few cell layers of a vessel (Krogh 1919). In addition to this, the process of tumour neovascularization often produces aberrant or leaky vessels which further Chapter 1 - Introduction 3 complicates the situation (Peterson and Appelgren 1977; Dvorak et al. 1988; Hashizume et al. 2000; Dvorak 2002). It is these limitations in diffusion and perfusion within solid tumours and resultant variable tissue oxygenation and acidification that define the tumour microenvironment (Sutherland 1988; Jain 1994; Tannock 2001; Minchinton and Tannock 2006). ARCHITECTURE Examples of two tumour architectures commonly observed in experimental tumours grown in mice are shown in FIGURE 1-1. Specimens were multiply immunostained to reveal the relation between vasculature, proliferation, hypoxia and necrosis. In both examples the tumours are fast growing, subcutaneous tumours in which the majority of tissue is comprised of tumour cells. Panel A shows a sparsely vascularized HCT-116 human tumour xenograft in which cords of tumour cells can be seen to surround vessels. In this example clear gradients in tissue oxygena-tion and proliferation are seen in relation to vasculature. Most cells are within ~150 um of a perfused vessel beyond which hypoxia is seen to occur, shown in green, followed by necrosis. Panel B shows an example of the experimental SCCVII mouse tumour which exhibits a dense vasculature, where most tumour tissue is within ~50 um of a vessel. In this case the hypoxic areas do not appear to be the product of diffusion limited oxygen and more likely arise from intermit-tent blood flow (Chaplin et al. 1987). Proliferating cells can be seen in both the green hypoxic areas and in the vicinity of unperfused blood vessels, consistent with the regular opening and closing of vessels. Chapter 1 - Introduction 4 •* . . . . . . • .fc.3w. - : — 7 - 1 A * t * + 150 j i m perfusion marker (DiOC?) • non-perfused vessels FIGURE 1-1 Examples of two tumour architectures commonly observed in experimental tu-mours grown subcutaneously in mice. A | A human HCT-116 tumour xenograft, which exhibits a sparse vasculature where cords of tumour cells are often seen to surround individual vessels. B | An SCCVII mouse tumour, which exhibits a dense vasculature and significant intermittent flow. Source: Minchinton Lab. Chapter 1 - Introduction 5 A n example of a clinical human breast cancer is shown in FIGURE 1-2. The section, stained using haematoxylin and eosin, shows nests of tumour cells existing within the normal tissue stroma. Blood vessels, observed by the presence of red haemoglobin, are seen in stromal areas but not within the tumour regions. In this case the relation between vasculature, oxygenation and proliferation w i l l be more complex with perhaps even greater diffusion distances involved. wmBmmmmf ,y . • • , . i .»#> J A , *«!!'*• Ah* *** « FIGURE 1-2 Example of a clinical human breast cancer. Three distinct regions are seen: i | vascularized stroma, arrows indicated visible haemoglobin, i i | unvascularized tumour cell nest and ii i | areas of necrosis. Photomicrograph of core biopsy courtesy of D r Malcom Hayes, pathol-ogy department B C Cancer Agency. FACTORS THAT DEFINE T H E TUMOUR MICROENVIRONMENT Diffusion limitations The sparse or aberrant vasculature that is often observed in solid tumours leads to diffusion limited gradients of potentially any molecule supplied and removed by the blood. Early evidence of a diffusion l imit for oxygen supplied from the blood was obtained from observations that their often appeared to be a cut off, in the range of ~150 um, to which viable tumour cells were observed, beyond which occurred necrosis (Thomlinson and Gray 1955). The hypothesis of the existence of oxygen gradients i n tumour tissue was later confirmed using radiolabelled misoni-dazole, which is reduced and forms adducts under conditions of low oxygenation (Chapman et al. 1981; Urtasun et al. 1986). Oxygen gradients are now easily detected using antibodies against several molecules (e.g. pimonidazole, EF5), which are reduced and bound under conditions of low tissue oxygenation (Raleigh et al. 1987). Limited diffusion also applies to other important Chapter 1 - Introduction 6 molecules such as glucose (Groebe et al. 1994; Venkatasubramanian et al. 2006) and the metabolic waste product lactic acid, build-up of which has been shown to be responsible for increasing the acidity of tumours (Tannock and Rotin 1989; Gatenby and Gillies 2004). All three of these factors, oxygen, glucose and lactate can on their own effect a reduction in proliferation of tumour cells. Hence it is not surprising that there is also a reduction in proliferation away from blood vessels (Tannock 1968; Hirst and Denekamp 1979). Variable vessel perfusion Variable blood vessel perfusion was originally demonstrated in experimental mouse tumours grown using SCCVII cells, where it was found that at any given time on the order of 20% of the vessels exhibited intermittent flow (Chaplin et al. 1987). Redundancy of vessels and short periods of closure may explain why tumour cells are able to maintain division under these conditions, see FIGURE 1-1B. In the original study, evaluation of blood flow status was performed through mismatch experiments in which two different fluorescent markers of vessel perfusion were injected intra-venously at 20-minute intervals (Chaplin et al. 1987). Mismatch of the two dyes seen in histological tissue sections was then used to determine the fraction of intermittently perfused vessels. Evaluation of the duration of closure and re-opening of vessels has been more recently examined using a tumour window model (Kimura et al. 1996). In this assay, a small glass window is attached to the skin of mice and tumours are grown against the window. The passage of blood cells through the vessels adjacent to the window can then be recorded via video microscopy. Results from such experiments have suggested closure and opening can occur repeatedly with a cycle time as short as tens of minutes (Dewhirst et al. 1998). Harsh conditions & aggressive phenotypes Diffusion limited regions of chronic reduced oxygen and glucose and elevated acidity that exist in tumours are compounded by irregular flow which can lead to cyclical re-oxygenation in some tumour regions, potentially creating an environment of repeated oxidative stress leading to DNA damage and mutation (Kimura et al. 1996). In addition, leaky vessels in tumours allow in mole-cules that would not normally have regular access to extravascular space (Peterson and Appelgren 1977; Dvorak et al. 1988; Hashizume et al. 2000; Dvorak 2002). All this creates a wide range of environmental conditions which, when linked with genomic instability, may be respon-sible for producing aggressive phenotypes seen in later stages of tumour progression (Harris 2002). In fact, the gradients in physiological parameters such as oxygenation, acidity and prolif-eration that exist away from vasculature that are observed in solid tumours could be considered the ideal system for production and selection of cell variants that will ultimately lead to tumour progression (Gatenby and Gillies 2004; Gatenby et al. 2006; Smallbone et al. 2006). Chapter 1 - Introduction 7 MODULATION OF THERAPY The tumour microenvironment, its modulation of anti-cancer therapy and its effect on tumour cell progression towards malignancy are topics currentiy under intense research. Perhaps the earliest observation of a potential for modulation of cancer therapy by the tumour microenvi-ronment was made by Thomlinson and Gray in 1955. Upon observing sparse tumour vasculatures, as depicted in FIGURE 1-2, in which blood vessels feeding tumour cells were con-fined to stromal tissue, they hypothesised that reduced tissue oxygenation might exist in the tumour regions most distant from vasculature, which would result in the existence of radiation resistant subpopulations (Thomlinson and Gray 1955). Oxygen . . ... -p MS& 1 1 ~5% 0 2 near vessels down to <0.5% far from vessels [^ physiological'effect] reduced proliferation, increased glucose reliance, increased lactate production [ ' effect on therapy ] reduced radiosensitivity, reduced effect of some drugs, e.g. bleomycin, etoposide counter rneagMrej repeated treatments, radio sensitizers (e.g. nimorazole), hypoxic cytotoxins (e.g. tirapazamine) Proliferation .gradient j u p (.0 ^()f0 c e \ \ s [n S-phase near vessels, down to 0% far from vessels'3 effect on therapy } reduced effect of cell cycle dependent drugs [ " c o u n t e r measurej repeated treatments, continuous infusions, use of drug combinations Acidification - . . , B r j d , l e n t gradient not known, extracellular pH is likely ~7.5 near vessels down to -6.8 far from vessels0 ^.phy^siological.effectl degradation of extracellular matrix, increased invasion, mutagenic effect on thorapy modified cellular uptake of many drugs (e.g. doxorubicin), radioresistance counter measure j repeated treatments, use of drug combinations Drugs Y'':-: ',: }vUS:Sffffien't] free plasma levels near vessels down to potentially neglegible levels far from vessels I p h y s ^ l o g i ^ l ^ f f e c t j potential for cultivation of drug resistance mechanisms [ clfect oh therapy] reduced exposure, increased cell survival ' counter measure j repeated treaments, increase exposure TABLE 1-1 Diffusion/perfusion limited gradients in solid tumours and their effect on therapy. a(Thomlinson and Gray 1955) (Raleigh et al. 1987), b(Tannock 1968; Hirst and Denekamp 1979), c(Tannock and Rotin 1989; Gatenby and Gillies 2004) Successful strategies for anti-cancer therapy must be able to address a wide range of environ-mental factors that will modulate treatment efficacy. TABLE 1-1 summarizes gradients that are commonly observed in solid tumours. Gradients in tissue oxygenation likely reduce the efficacy of radiation (Thomlinson and Gray 1955; Moulder and Rockwell 1987; Janssen et al. 2005) as well as drugs such as bleomycin and etoposide (Yamauchi et al. 1987). To attempt to counter this, oxygen mimetic radiosensitjzers, which substitute for oxygen to prevent reconnecting of radia-tion induced DNA strand breaks, were developed e.g. misonidazole and nimorazole (Overgaard et al. 1989; Overgaard et al. 1998). Hypoxic cytotoxins, which are enzymatically converted to cytotoxic species under conditions of low oxygen, were later developed in an attempt to exploit Chapter 1 - Introduction 8 AQ4N (Brown 1993; Patterson and McKeown 2000). Increased tumour acidity modifies the activity of many anticancer agents (Tannock and Rotin 1989), for example, reducing the cellular uptake and efficacy of weak bases such as the anthracyclines (Mahoney et al. 2003). Most anti-cancer drugs in current use were developed to exploit the increased proliferation status of tumour cells, hence reduced cellular proliferation modulates the efficacy of the majority of anticancer drugs in existence. Finally, diffusion limited penetration can result in lower drug exposure in the tumour regions that are distant from vasculature, resulting in reduced efficacy. The repeated exposure to sub-toxic drug levels is perhaps the simplest way of cultivating a drug resistant phenotype. In theory the gradients in oxygen, acidity and drugs that can occur in solid tumours will effect changes in treatment response. However direct causal evidence linking these problems with clinical management of cancer is often lacking. 1.2 DRUG PENETRATION In contrast to most normal tissues, with relatively high microvessel density, the extravascular compartment of solid tumours may pose a significant barrier to the penetration of molecules supplied by the blood. However, little is known of the ability of many of the most commonly used anti-cancer drugs to distribute within solid tumours (Tannock et al. 2002). Measurement of gross tumour drug accumulation is a routine component of drug development but it tells little about extravascular drug distribution. This is especially true in the case of drugs that are highly reactive in tissue, which may produce high tumour accumulation that is confined to the cells immediately surrounding active vessels. In principle, visualization of a drug using radiolabelled or immunohistochemical-based assays is the most direct way to evaluate drug penetration. However, this approach is not always possible and has limited applicability as a fast, generic screening assay for newly synthesised drugs, especially when stability issues/metabolites are not known. FACTORS INFLUENCING TUMOUR DRUG DISTRIBUTION: SUPPLY A N D D E M A N D Like the gradients in oxygen that were first postulated by Thomlinson and Gray (Thomlinson and Gray 1955), the tumour distribution of any molecule supplied from the blood will be subject to the constraints of supply and demand. TABLE 1-2 lists factors that can play important roles in determining the tumour distribution of a drug. Supply factors are controlled by pharmacokinetic variables such as dose, metabolism, excretion and blood binding. Flux of drugs within tumour tissue is controlled by factors such as MW, lipophilicity, charge etc.. Factors that determine demand of the drugs by tissue can include cellular metabolism, binding and sequestration. Chapter 1 - Introduction Pharmacokinetics dose metabolism' determines C m a x / A U C , T 1 / 2 I'M n'lioii , , . ,. will limit free drug available for diffusion, may serve as a reservoir of drug that is slowly oo uii mg. | . l , | i , a s e c j w h i i e protected from metabolism in liver / issi/i- Clumictcris-tics often low in tumours, increases gradients of diffusion limited molecules \J--IUI.1I Jl'llsllV a b typical max. distance to a vessel: normal tissue 3-5 cell layers ; tumours up to 15-20 cell layers vascular integrity often compromised in tumours, increases access to extravascular space interstitial pressure often elevated in tumours may lead to intermittant blood flowd will determine traction or avauaoie volume tor diffusion tor drugs mat enter cells siowiy extracellular fraction c tvrjical extracellular fraction observed in tumours: 35-60% extracellular matrix charged matrix that may modify the diffusion of some drugs6 Drug Characteristics MIV inversely proportional to a molecule's diffusion coefficient in water lipuplulii llv -vvill control access of drugs to cellular compartment, easy access means more flux through charge n s s u e but also possibly more cellular consumption Drug-cell interactions nu tabol ism consumption of a drug may limit its tissue penetration lundm,: leads to accumulation of drugs, may limit tissue penetration, may serve as a reservoir that is buquesteri/ation slowly released after drug is cleared from plasma I*np status will reduce access to cellular compartment, may increase tissue penetration TABLE 1-2 Factors which wi l l influence a drug's ability to penetrate and distribute within solid tumours. a (Krogh 1919), b(Thomlinson and Gray 1955), c(Peterson and Appelgren 1977; Dvorak et al. 1988; Hashizume et al. 2000; Dvorak 2002), d (Chaplin et al. 1987) (Kimura et al. 1996) (Jain 1987; Dewhirst et al. 1998), "Gain 1987). Upon leaving the vasculature and entering the extravascular tumour compartment a drug's ability to penetrate and distribute within tissue will depend on the balance between its ability to diffuse into tissue versus its rate of sequestration/consumption by the tissue, FIGURE 1-3 illus-trates potential drug-tissue interactions that may limit tissue penetration for a hypothetical anthracycline-like drug. In panel A , the drug is seen to either be confined predominantly to the extracellular space, with occasional entry and binding within cells (upper path) or to have easy access to the cellular compartment (lower path). In the second case the increased sequestra-tion/consumption of the drug due to its greater access to the cellular compartment may or may not limit its tissue penetration. Panel B illustrates the known drug-tissue interactions of doxorubi-cin. Interactions include D N A intercalation (Zunino et al. 1980), sequestration in acidic lysosomes (Coley et al. 1993; Hurwitz et al. 1997), passive uptake in proportion to lipophilicity and un-charged fraction and active efflux (Frezard and Garnier-Suillerot 1991) and membrane binding (de Wolf et al. 1993; Demant and Friche 1998a). Potential extracellular interactions include negatively charged extracellular matrix components as well as membrane and protein binding (Chassany et al. 1994). The mathematical modelling of any drug's tissue penetration will require detailed knowledge of its physicochemical properties and all significant cellular interactions. Chapter 1 - Introduction 10 FIGURE 1-3 A | Examples of two possible routes for a hypothetical anthracycline-type drug that accumulates through DNA intercalation. The upper path depicts the route if the drug is mostly confined to the extracellular space and the lower path depicts the route if the drug can easily access the cellular compartment. B | Schematic of known cellular interactions of the anthracy-cline doxorubicin: l | diffusion, 21 passive cellular uptake/active efflux, 3|DNA intercalation, 41 acidic sequestration, 51 interactions with extracellular proteins and charged components of the extracellular space, 6 | phospholipid interactions. Adapted from (Kyle 1999). TUMOUR MAPPING Recent advances in immunohistochemistry and digital imaging mean that many drugs can now be visualized and mapped in relation to tumour vasculature. Key innovations were the develop-ment of antibodies to detect vasculature, e.g. anti-CD31, Von Willebrand factor VIII, improvements in quantitative digital imaging, robotic microscopy and computer processing. For the last 5-10 years it has been possible to capture images of entire tumour sections at high enough resolution to map drug levels in tissue relative to tumour vasculature. Reports are now beginning to appear that examine radiation and drug effects in relation to immunodetected tumour vascula-ture. Examples include doxorubicin as detected by fluorescence (Primeau et al. 2005), radiation via effects on proliferation (Ljungkvist et al. 2006) and gemcitabine via effects on proliferation (Huxham et al. 2004). Direct imaging of drugs and their adducts Studies examining drug distribution without immunodetected tumour vasculature information have looked at radiolabelled and fluorescent drugs and dyes. The earliest studies carried out looked at the diffusion of lissamine green and other dyes in tumours (Goldacre and Sylven 1962). Doxorubicin is perhaps the most studied anti-cancer drug due to its fluorescence (Ozols et al. 1979; Lankelma et al. 1999), though in general no attempt has been made to differentiate it from its less active but fluorescent metabolite doxorubicinol or between free and partially quenched bound drug. Radiolabelled drugs studied include paclitaxel (Li et al. 2000) and immunotoxins (Sung et al. 1993). In addition to classical detection via radiolabelling or direct fluorescence there now exist a number of drug adducts which can be detected via antibody recognition. Examples of drugs that can be detected by antibodies include cisplatin-DNA adducts (Terheggen et al. 1987; Tilby et al. 1991; Liedert et al. 2006). The thymidine analogue, Bromodeoxyuridine (BrdUrd), Chapter 1 - Introduction 11 originally developed as a radiation sensitizer is incorporated into DNA during replication and can be detected via antibodies raised against DNA-BrdUrd adducts (Gratzner 1982). Doxorubicin has also been reported to have been detected immunohistochemically (Henneberry and Aherne 1992). Other examples include immunodetection of adducts formed by the 2-nitoimidazoles such as EF5 and pimonidazole binding to thiol groups in proteins and amino acids under hypoxic conditions (Raleigh et al. 1987; Lord et al. 1993; Azuma et al. 1997). Mapping drug effect in tumours Markers for proliferation (e.g. BrdUrd, Ki67, PCNA), apoptosis (e.g. Tunel, caspase 3) and double strand breaks (e.g. yH2AX) can also be used in an attempt to map drug distributions via their effect on cells in tumours relative to vasculature. However, the relation between drug effect and drug distribution is generally obscured by changes in the intrinsic sensitivity of cells to drug with depth into tissue. Hence these studies address more general questions of drug efficacy in solid tumours rather than questions specific to drug penetration. The earliest mapping studies were done in spheroids and tumours using either sequential trypsinization or cell-sorting techniques with cell survival as the endpoint (Giesbrecht et al. 1981; Durand 1982; Olive et al. 1985). For the cell-sorting method, treated cells were dissociated and sorted based on the strength of Hoechst 33342 fluorescence, which was allowed to diffuse into the tissue and used as an indicator of the distance of each cell from a perfused vessel/tissue edge. Work was first carried out with radia-tion and doxorubicin (Chaplin et al. 1985; Durand 1990) and later other anti-cancer agents (Olive and Durand 1994). New approaches involve imaging of tissue sections for markers of vasculature and drug effect. Using the tissue section-based techniques, the effects of radiation (Ljungkvist et al. 2006) and drugs such as gemcitabine (Huxham et al. 2004) and tirapazamine (Huxham et al. 2006) have been evaluated in relation to tumour vasculature. Limitations of tumour mapping for evaluation of drug penetration While histological based tumour mapping represents a relatively new and unexplored area of research, numerous limitations exist in terms of employing it in assessing drug penetration. With regards to direct imaging, radiolabelled drugs are not always available and are generally expen-sive. Autoradiography-based experiments are labour intensive and getting results can entail waiting weeks or months while film is being exposed. In addition, the potential for presence of radiolabelled drug metabolites means that data will often be inconclusive. Assays involving biological endpoints are prone to changes in intrinsic sensitivity of cells to drugs with changing proliferation status, oxygenation etc. (see TABLE l-l). In addition, indirect drug-derived host-toxicity may result in changes in tumour proliferation and even apoptosis via a reduction in oxygen and nutrients supplied from the blood. Chapter 1 - Introduction 12 1.3 IN VITRO TUMOUR MODELS There exist several in vitro models for evaluating tumour drug diffusion, the best characterized and most widely used being multicellular spheroids (Sutherland 1988). Other models include the one used here, multilayered cell culture (Cowan et al. 1996; Minchinton et al. 1997). In addition, chunks and slabs of tissue excised from solid tumours and cellular aggregates grown on tissue culture plates or embedded in collagen or sponge are also used (Leighton 1951; Swabb et al. 1974; Yang et al. 1979; Vescio et al. 1987). SPHEROIDAL CELL CULTURE Multicellular spheroidal cell culture is the mainstay of in vitro 3-dimensional tumour modelling. Developed over 30 years ago in the Sutherland Lab at the University of Western Ontario, sphe-roids have been shown to reproduce many of the features of the tumour microenvironment. Diffusion limited penetration of molecules into the cultures means they can reproduce the oxygen and nutrient gradients and gradients in proliferation seen in solid tumours. Thicker spheroids exhibit acidified central regions, due to lactate build-up, and if large enough they form a central necrotic region. In addition they have been shown to exhibit a 3-D culture cell contact effect which modifies their response to radiation and drugs. Because they are able to integrate a number of characteristics of the tumour microenvironment they are an ideal system for model-ling the in vivo response to anti-cancer agents. Modelling solid tumours Spheroids have been applied to study relations between oxygen, glucose, serum and prolifera-tion/necrosis (Freyer and Sutherland 1983; Freyer et al. 1984; Casciari et al. 1988; Acker et al. 1992; Groebe and Mueller-Klieser 1996). Diffusion limited penetration of molecules into the cultures means that as well as producing oxygen and nutrient gradients they can also develop an acidic central region due to lactate accumulation (Walenta et al. 2000) and have been used to study relations between central pH and medium buffering capacity (Acker et al. 1987). They have been grown using a number of cell types including normal and drug resistant variants (Martin et al. 2003) and have been shown to create an extracellular matrix (Glimelius et al. 1988). Effect of tumour microenvironment on response. Cells located on the outer versus inner layers of spheroids have been shown to exhibit differential sensitivity to a number of drugs and also to radiation (Sutherland et al. 1970; Durand and Suther-land 1972; Sutherland and Durand 1976; Durand 1980; Olive and Durand 1994). Changes in sensitivity of cells have been attributed to a number a factors including cell cycle, inter-cellular communication and drug penetration. Chapter 1 - Introduction 13 Drug penetration The tissue penetration of drugs from several classes has been studied using spheroidal cell culture. Direct visualization of the anthracycline doxorubicin (Kerr et al. 1988) and radiolabelled vinblastine (Nederman et al. 1981), methotrexate (West et al. 1980) and other biological agents (Nederman et al. 1988) have been investigated using spheroids. Indirect effect-based detection has been used to study several drugs including doxorubicin, cisplatin and others (Durand 1990; Olive and Durand 1994). However, in these studies it was shown that the differential response of inner versus outer cells in spheroids depended on a number of factors. MULTILAYERED CELL CULTURE Multilayered cell culture (MCC) is a variation on spheroidal cell culture in which cells are grown as discs of tissue. Cultures are grown starting from a single cell suspension which is added to a tissue culture insert and allowed to attach overnight to form a confluent layer. Inserts are then transferred to stirred reservoirs and grown for 2-5 days to form planar cultures typically 15-30 cell layers thick, see FIGURE 1-4. A & s X H B H a i H H H B 1-1 FIGURE 1-4 Multilayered cell culture (MCC) growth apparatus and histology. Cultures are grown using standard tissue culture inserts (Millipore CM). A | Cells are first seeded and allowed to attach to the porous PTFE membranes pre-coated with collagen. B | Cultures are then im-mersed in stirred growth medium where they have access to oxygen and nutrients from both sides. C | Haematoxylin stained MCC cryosection showing ~15-20 cell layers resting on the PTFE membrane. MCC-based flux assay MCC was developed to allow measurement of flux of drugs through tumour-like tissue in a controlled environment. There are now numerous studies published using this application for a variety of anti-cancer drugs including radiosensitizers (Cowan et al. 1996; Kyle 1999), hypoxic cytotoxins (Hicks et al. 1998; Phillips et al. 1998; Kyle and Minchinton 1999), DNA-intercalators (Hicks et al. 1997) and other agents (Tannock et al. 2002). FIGURE 1-5 shows schematics of single Chapter 1 - Introduction 14 and a dual-reservoir conformation that are often used for measurement of drug flux. The dual-reservoir configuration allows drug concentration from both donating and receiving reservoirs to be monitored and allow modelling of mass balance. FIGURE 1-5 C-D show typical data produced using the dual-reservoir configuration. Fitting both data sets simultaneously allows for mainte-nance of mass balance between reservoirs and monitoring of net drug sequestration and consumption in the culture, FIGURE 1-6 illustrates the basic process of drug build-up within the cultures that occurs during flux experiments. Flux-based assay 0 1 0 20 30 40 50 Time (hours) FIGURE 1-5 Schematics of two types of apparatus used for MCC-based flux experiments and typical data produced by the assay. A | Single reservoir with floating MCC and BJdual reservoir configuration. In A, the unstirred MCC is filled with agar, premixed with the drug to prevent convection; samples are then taken from the stirred reservoir over time. In B, drug is added to one side and the flux through the culture and into the second reservoir is monitored over time. C-D | Typical data from donating and receiving reservoirs using the dual-reservoir configuration. In this case drug diffuses through a tissue culture insert membrane with no cells. In C, the drug misonidazole diffuses without loss and drug disappearance from the donating reservoir is matched by its appearance in the receiving reservoir. In D, doxorubicin is lost through instability in the growth medium and the change in the donating reservoir is not matched by appearance in the receiving reservoir. Adapted from (Kyle 1999). Results from flux experiments can be used for direct comparison of relative flux between two or more drugs or experimental conditions or used to estimate rates of diffusion and sequestra-tion/consumption within the tissue. On their own, raw data from one or both reservoirs can be used to compare the relative flux of two or more drugs. It is also possible to examine the effect of Chapter 1 - Introduction 15 altering some experimental parameter, often the inhibition of a specific consumptive process. Examples of this include variation of tissue oxygenation to inhibit the enzymatic conversion of the hypoxic cytotoxin tirapazamine (Hicks et al. 1998; Phillips et al. 1998; Kyle and Minchinton 1999) and over-expression of Pgp to reduce cellular accumulation of doxorubicin (Tunggal et al. 2000) . Data from flux experiments can also be fitted to differential equations to obtain estimates of parameters representing diffusion and sequestration/consumption processes within the cultures. Examples include analysis of the DNA intercalator DAPA (Hicks et al. 1997), the hypoxic cytotoxin tirapazamine (Hicks et al. 1998; Kyle and Minchinton 1999) and recently a panel of tirapazamine analogues (Hicks et al. 2006). Flux assay: process of drug build up Pre-s teady state: transient |uild ;up, Co' Steady state: simple diffusion: linear gradient ! • • •» diffusion+loss:non-linear gradient Co FIGURE 1-6 Schematic of the process of drug build-up within an MCC during flux experiments using the dual-reservoir configuration. Following addition of drug to one reservoir, a phase of transient build-up occurs within the culture. During the subsequent steady-state phase, for drugs that are not sequestered/consumed in the tissue a linear gradient will form within the tissue and the amount of drug entering the culture from the donating reservoir will be matched by drug exiting the culture to the receiving reservoir. If loss does occur within the tissue, a non-linear gradient will link concentrations in the two reservoirs and less drug will exit the MCC than enter it (Crank 1975). Chapter 1 - Introduction 16 Simulating drug distribution In cases where the key drug-tissue interactions are understood and have been quantified it may be possible to mathematically model the tissue pharmacokinetics of a drug. Once the initial characterization work has been carried out this approach provides flexibility in terms of explor-ing the effect of variations in exposure conditions such as plasma pharmacokinetics or tumour architecture on drug distribution (Kyle and Minchinton 1999; Hicks et al. 2006). However, problems may be encountered if this approach is used to model an under characterized system or through oversimplification of complex drug-tissue interactions, potentially leading to false predictions. FIGURE 1-7 Apparatus used to expose M C C to drug from just one side. A | MCCs are placed in a stirred reservoir with the membrane side of each tissue culture insert clamped against the reservoir wall. The culture then has access to growth medium and drug from one side only, thereby allowing build-up of drug to occur within the tissue. B | Following drug penetration experiments cultures, are embedded in cryosectioning gel (O.C.T.) and frozen with the bottom of the tissue culture insert in contact with an aluminium block. Visualizing drug fluorescence in MCC An alternative application of MCC to assess the distribution of drugs is to directly examine drug within the tissue. One way of doing this is to close-off one side of the MCCs and allow drug to build-up from the other side, as shown in FIGURE 1-7. Examples of the possible range of variation in drug profiles within 150 jum thick cultures follow-ing a 1-hour drug exposure to one side are shown in FIGURE 1-8. Panels show drug-derived fluorescence visualized in tissue cryosections. Fluorescent profiles indicate that in some cases drug can be limited to just the first few cell layers, while in others, drugs can exhibit almost uniform distribution over the full width of the tissue. Of the examples shown here the fluores-cent-tagged goat IgG, with a molecular weight over 100 times greater than the other molecules, not surprisingly exhibits the poorest distribution after the 1-hour exposure. However, ranking of the other agents doesn't strictly follow molecular weight. Factors that can modify each drug's tissue penetration will include their level of access to the cellular compartment and degree of sequestration/consumption within the tissue. Comparison of two the Hoechst drugs examined here indicates that 33258 exhibits a more uniform tissue distribution than 33342 despite 33342 being more lipophilic and entering cells A B Chapter 1 - Introduction 17 more easily than 33258. Both drugs accumulate in cells through binding to DNA. In the case where a drug's accumulation is negligible in comparison with its rate of flux through tissue, a more lipophilic drug will have better access to the intra-cellular compartment of the tissue and hence exhibit faster flux. However, if access to the intra-cellular compartment leads to significant accumulation then the opposite may occur and the less lipophilic drug, confined to the extra-cellular compartment, will exhibit a more uniform distribution. In this case it is 33258, the less lipophilic drug, which exhibits the more uniform distribution, suggesting that drug sequestration via DNA binding may be high enough to limit its tissue penetration. Fluorescence-based evaluation of drug penetration. 1 0.8 0.6 0.4 0.2 0 Goat IgG - 4 0 /jg/ml MW 7 5 000 Hoechst 33342 - 5 fjM MW 562 DiOC7(3) -1 /JM MW600 Rhodamine 6 G - 0.1 / / M M W 4 7 9 Irinotecan - 1 0 0 ^/M M W 5 8 7 Hoechst 3 3 2 5 8 - 5 / J M M W 538 0.8 0.6 0.4 0.2 Tetracycline - 100f/M M W 4 4 4 Topotecan - 1 0 / J M M W 4 2 1 Fluorescein - 1 0 £/M M W 3 3 2 50 100 150 0 50 100 150 0 Depth into tissue (/jm) 50 100 150 FIGURE 1-8 Drug derived fluorescent profiles observed in HCT-116 M C C s following 1-hour exposure using the closed-off apparatus. Drug profiles are not seen to rank solely by MW. Following drug exposure cultures were embedded in cryosectioning medium and immediately frozen. Panels show fluorescence as a function of depth into tissue as determined from epifluo-rescence imaging of unmounted dry-cryosections. Hoechst 33342, hoechst 33258 (excitation 365, emission 460 nm); irinotecan, topotecan, tetracycline (ex 425 nm em 540 nm); DiOC7(3), fluo-rescein (ex 480 nm, em 535 nm); goat IgG -Alexa 546, rhodamine 6G (ex 540 nm, em 605 nm). Source: Kyle unpublished. Chapter 1 - Introduction 18 Fluorescein and topotecan, which appear to distribute the most efficiently of the agents studied, both exhibit oddly shaped profiles. This is likely due to an unusually fast washout during the freezing process, consistent with a molecule that exhibits rapid diffusion and negligible retention through cellular accumulation or binding. During freezing, the edges of the cultures come into contact with embedding gel for 1-2 min before it solidifies, FIGURE 1-7. Freezing occurs from the bottom /far edge of the cultures, which likely explains why there appears to be less washout from that side. 1.4 PROJECT OUTLINE: DEVELOPING A NOVEL M C C - B A S E D ASSAY FOR ASSESSING D R U G DISTRIBUTION USING A BIOLOGICAL ENDPOINT The objective of this research project was to develop and evaluate a novel MCC-based assay that would measure the penetration and distribution of drugs in tissue using a biological endpoint. A unique property of the disc geometry of MCC is that it presents a barrier to penetration in which there exist two populations of rapidly proliferating cells, one on each edge of the culture, that are separated by a known thickness. This symmetry presents an opportunity to use the effect of the drugs on proliferating cells as an indication of their penetration, in effect using the cells them-selves as the drug detection end-point. In tumours and spheroids this technique is not possible because proliferation and cellular response fall off with depth into tissue and hence drug penetra-tion is only one of several factors that will determine cellular response. EFFECT-BASED DRUG DISTRIBUTION ASSAY Like spheroids, MCCs exhibit a gradient in proliferation but since they grow as planar discs, this gradient forms as a mirror image from either side towards the middle, FIGURE 1-9. In this study, we exploit this symmetry by exposing cultures to drugs from one side and then comparing their effect on the cells located on the near (exposed) side versus the far side of the cultures. FIGURE 1-9 Symmetrical S-phase cell distribution seen in an H C T - 1 1 6 M C C as detected by immunohistochemical staining of BrdUrd incorporated into D N A . Image of M C C cryosection, 10 um thick, showing immunodetected BrdUrd and counterstained with haematoxylin. BrdUrd exposure and staining were carried out as described in section 2.3. Choice of boundary conditions Drug exposure can be made with the far side of the cultures either open or closed-off to growth medium, the in vivo analogue of each of these configurations is depicted in FIGURE 1-10. Expo-sures made with the far side of the cultures closed-off, FIGURE 1-7, represent the best-case scenario Chapter 1 - Introduction 19 in terms of allowing drug to build-up within tissue and are most closely related to the in vivo situation in which there exist multiple blood vessels feeding a region of viable tissue, as depicted in FIGURE 1-10A. A second "open" situation, where an MCC is placed between two reservoirs, of which one contains drug, would represent a greater barrier to achieving a uniform distribution of drug, and most closely resembles, but may exceed, the in vivo situation where viable tissue is interspersed with necrosis, FIGURE 1-10B. FIGURE 1-10 Examples of two possible boundary situations: closed-off versus open. A | Symmetrical diffusion outwards from two blood vessels leads to build-up of drug at the mid-point. This configuration, approximated by the closed-off MCC, represents the best-case scenario in terms of allowing drug to build-up within the tissue. B | Diffusion out from a single vessel beyond which occurs necrosis. In this case drug is not permitted to build-up at the tissue bound-ary but washes out into the necrotic region. This situation most closely resembles the open MCC configuration and poses a greater barrier to achieving a uniform distribution of drug within the tissue. Procedure An effect-based assay that could be used to determine a drug's distribution is outlined in FIGURE 1-11. After cultures are grown to their desired thickness they are placed in the drug exposure apparatus. The "closed-off" configuration was chosen for performing drug exposures to allow for the best-case scenario in terms of achieving a uniform distribution of drug. Using this configura-tion the maximum period of drug exposure will be limited to the duration that the cultures can be closed-off and still maintain symmetrical proliferation, likely of the order of hours. Following drug exposure, cultures will then be returned to the growth chambers and incubated for a period of time, of the order of days, to allow drug effect to manifest itself. During this incubation period drug may also redistribute and penetrate further into the culture. Boundary conditions during the incubation period match the "open" configuration to ensure that symmetrical proliferation is sustained. The incubation period must be long enough to allow manifestation of drug effect but short enough to avoid substantial re-growth of the cultures that would further complicate interpretation of the data. Following the incubation period the S-phase cell marker BrdUrd will be used to detect drug effect. For drugs that diffuse in and wash out of tissue quickly, it may be critical that the proliferation status of the cells on the far side of the cultures is not perturbed during the exposure period. Chapter 1 - Introduction 20 However in other cases the proliferation status on the far side of the cultures during drug exposure may not be as important. For example if the drug is significantly retained following the exposure period, then it may only be important that the cells are in proliferation during the incubation period. Alternatively, for drugs that accumulate and are retained in the first few layers of cells and later diffuse slowly into the tissue, significant exposure to the far side of the cultures may not actually occur until the incubation period, when the drug that accumulated in the first few layers of cells slowly redistributes within the tissue. In this case, it might not even be necessary to close off the cultures during the exposure period. Effect-based drug penetration screening assay Grow MCCs ifciUfcltti Drug Exposure 3 Incubate FIGURE 1-11 Outline of the assay proposed to evaluate drug distribution using the symmetrical proliferation observed in MCC. Once cultures are grown to the desired thickness they would be transferred temporarily to a penetration apparatus where one side would be closed-off and drug exposure take place. Cultures would then be returned to the growth apparatus and incubated for a period of time to allow for the manifestation of drug effect. Following this incubation period, cultures would then be assayed to detect differential effect to either side, in effect using the cells themselves as the drug detection device. Interpreting results Comparison of dose response curves from the exposed versus far sides of treated MCCs should allow for evaluation of drug distribution within the tissue. FIGURE 1-12 shows examples of some of the potential situations that may arise when the assay is applied to assess drug penetration. In the simplest case, FIGURE 1-12B, the drug exposure to the far side of MCCs will be proportional to the exposed side. In this case the relative distribution of drug will be concentration independent and the reduction in exposure to the far side can be determined by measuring the shift between the two dose response curves. In the case where drug distribution is concentration dependent, FIGURE 1-12C, as might occur when a saturable process is involved in limiting drug penetration, the gap between dose response curves may become smaller as the driving concentration is increased. Another situation may arise for drugs that transiently increase the fraction of prolifer-ating cells in the tissue or induce immediate cell loss at high enough exposures. Both these cases may actually produce an increase S-phase labelling at high drug exposures, FIGURE 1-12D. In the case of drug induced cell loss, the apparent increase in proliferation that might be observed would be because the cells located on the exposed edge at the end of the experiment would Chapter 1 - Introduction 21 actually have been located deeper into the tissue at time of drug exposure and hence shielded from the drug. exposed s ide far s ide log [drug] log [drug] log [drug] FIGURE 1-12 Estimating drug distribution based on comparison of dose response curves on exposed versus far sides of MCCs. A | Proliferation seen in an MCC following drug exposure from one side. Hypothetical dose response curves obtained from analysis of proliferation on exposed versus far sides of MCC following treatment for the following cases: B | drug distribution profile is concentration independent, C | drug distribution is concentration dependent and D | cell loss or cell cycle perturbation occurs at higher concentrations. PROJECT HYPOTHESIS The symmetrical proliferation seen in MCCs can be applied to measure the ability of anti-cancer drugs to penetrate and distribute within tissue using a biological endpoint. DEVELOPMENT A N D VALIDATION OF T H E MCC EFFECT-BASED ASSAY The development and validation of the assay will need to address the following questions: -What effect does temporarily shutting-off one side of the culture have? Can symmetrical prolif-eration be maintained during the temporary closure period during drug exposure. -What control assays can be performed to detect non-diffusion related differences in drug effect to the two sides? For instance, effects related to reduced oxygen and pH gradients that may form during the closure period. -Does the assay correlate with direct visualization of drug in MCCs? -How do results from MCC tissue correlate with tumour xenograft based work? -Is the S-phase cell marker BrdUrd a good end-point for the MCC based work? Can a reliable BrdUrd-based assay be developed for detection of drug effect that can be used in MCC and also in tumour xenografts grown in mice. Chapter 1 - Introduction 22 -What classes of drugs can be tested using the assay? -What are the limitations of the assay? 1.5 OUTLINE OF PRESENTED WORK SUITABILITY OF T H E S-PHASE MARKER BRDURD AS BIOLOGICAL END-POINT An initial decision was made to assess the S-phase marker BrdUrd as the end-point used to evaluate drug effect. Choices were between markers of cell death/apoptosis (TUNEL, activated caspase-3) and markers of proliferation (BrdUrd, Ki-67, PCNA). We chose to start by trying a marker of DNA synthesis due to the large existing body of knowledge of the cell cycle effects of most anticancer drugs derived from tritiated thymidine based work and more recently BrdUrd and flow cytometry work. Bromodeoxyuridine (BrdUrd) is a thymidine analogue that is incorporated into DNA during the S-phase of the cell cycle. Originally investigated as a radiation sensitizer (Djordjevic and Szybal-ski 1960; Kinsella et al. 1984b; Phillips et al. 1991; Ensminger et al. 1994), it has found use as a marker of cell proliferation particularly in cancer (Dolbeare et al. 1983; Begg et al. 1985; Wilson et al. 1985; Rew et al. 1992), where immunohistochemical detection of BrdUrd-DNA adducts (Gratzner 1982) makes it a simpler alternative to autoradiographic techniques using radiolabelled thymidine. Unlike endogenous markers for cellular proliferation such as PCNA or Ki-67, BrdUrd is exogenously administered and must reach proliferating cells via the vasculature. Upon leaving the vascular compartment, it then diffuses through cellular and extracellular compartments. In order for BrdUrd to act as an effective marker of cellular proliferation it must be able to distribute within the extravascular compartment with enough uniformity or at high enough concentration so that proliferating cells at any distance from a blood vessel are uniformly labelled. Plan of work To evaluate the feasibility of using BrdUrd as the effect-based assay end-point we set out to compare the ability of BrdUrd to penetrate and label proliferating cells in tumour xenografts and in MCCs. We began by looking at the dose dependence of BrdUrd labelled cells in relation to blood vessels in tumour xenografts and then carried out parallel experiments in MCCs. For the tumour based work the key objective was to determine if tissue penetration played a role in limiting BrdUrd labelling in cells located away from blood vessels and if so then what dose was required to avoid this phenomena. For the MCC-based work, in addition to determining the BrdUrd exposure required to label cells within the tissue, we also wanted to determine if prolif-eration could be sustained on the far side of the cultures during a short period of closure. Examination of BrdUrd exposure requirements in tumours and MCCs also provided an opportu-nity to compare the barrier to diffusion presented by the cells grown as tumour xenografts versus MCCs. Chapter 1 - Introduction 23 ANTHRACYCLINE DISTRIBUTION IN MCC: EFFECT-BASED ASSAY VERSUS DIRECT VISUALIZATION The anthracyclines, doxorubicin (DOX), epirubicin (EPI), daunorubicin (DAU) and the related compound mitoxantrone (MIT) were chosen as the first test group for the effect-based assay. A considerable body of knowledge surrounds their cellular interactions and there exist numerous reports on the limited ability of doxorubicin in particular to distribute within tissue. In spheroids, early work showed that doxorubicin exhibited limited tissue penetration (Sutherland and Durand 1976; Kerr et al. 1988). The potential for redistribution of the drug was shown in later studies where it was found that the initial distribution of cell kill following a 2-hour exposure could be improved if spheroids were then left intact for an additional 24 hours (Durand 1990). Work in experimental and human tumours has found that doxorubicin can exhibit gradients in relation to vasculature (Lankelma et al. 1999; Primeau et al. 2005). In addition to poor tissue distribution, further motivation for studying the anthracyclines was that the first three of these agents fluoresce hence allowing for a comparison to be made between the effect-based assay and direct visualization of drug fluorescence. While the drugs possess similar mechanisms of action their differing physicochemical properties and clinical activities suggest they may exhibit different tissue distributions. The agents are weak bases with pKa values of ~8.3 for DOX, DAU and MIT and ~8.1 for EPI (Di Marco et al. 1977; Mahoney et al. 2003) and at pH 7.4 are between 85-90% charged. Entry into cells is believed to occur via passive uptake of the neutral species for all four drugs. Of the three, DAU is the least hydrophilic, entering cells 3-10 times more quickly than DOX (Frezard and Garnier-Suillerot 1991; Demant and Friche 1998b), while EPI exhibits an uptake rate approximately twice that of DOX (Cantoni et al. 1990). Log octanol-water partition coefficients at pH 7.4 are DOX -0.3, EPI -0.6, DAU -0.9 and MIT -0.8 (Burns et al. 1988; Mulder et al. 1995; Demant and Friche 1998a). All bind readily to DNA and accumulate in cells (Frezard and Garnier-Suillerot 1990). From previous experience with these drugs we anticipate that the more rapid rate of cellular uptake of the more lipophilic drugs may compromise their ability to penetrate into tissue. Plan of work The work outline for this section was first to determine if doxorubicin produced a differential effect on proliferation on the two edges of the cultures following exposure to one side. As the first drug to be tested it was also necessary to determine conditions required to maintain proliferation on the far side of the cultures during drug exposure. Exposure duration, wait period to the BrdUrd end-point and method for data analysis needed to be determined. Assuming success with doxorubicin, the other three agents would be tested. Results from the effect-based assay would then be compared with direct visualization of the fluorescent agents under comparable conditions. As an additional benchmark, the rate of flux of the four agents through MCC was determined using the flux apparatus outlined in FIGURE 1-5B. Chapter 1 - Introduction 24 N H 2 DOXORUBICIN DAUNORUBICIN EPIRUBICIN R2 R3 R.i R OCH3 OCH3 OCH 3 H H OH OH OH H OH H OH OH O NHC^CHjNHCHjCH^H OH O NHCHjCHjNHCHjCHjOH MITOXANTRONE FIGURE 1-13 Chemical structures of the anthracyclines to be evaluated using the effect-based assay. Adapted from (Goodman et al. 2006) TAXANE DISTRIBUTION IN MCC & TUMOURS: EFFECT-BASED ASSAY & DIRECT VISUALIZATION As a second test group, we compared the tissue distribution of the taxanes, paclitaxel and docetaxel. Motivation for choosing the taxanes was based on published reports suggesting poor tissue penetration of paclitaxel and the commercial availability of radiolabelled versions of both agents, which enabled comparison of effect-based results from MCC with direct visualization in MCC and tumour xenografts. In addition it was hypothesised that the considerable differences in the physicochemical parameters between the two drugs might lead to differences in their tissue penetration. The tissue distribution of paclitaxel has been studied previously via detection of radiolabelled drug in whole tumour sections (Li et al. 2000), but without vasculature information, and also in tumour histoculfures (Kuh et al. 1999). These studies suggested poor penetration of paclitaxel though no analysis was performed. Paclitaxel tissue distribution has been studied quantitatively, via HPLC detection, in bladder tissue that was sequentially sectioned in relation to the urothe-lium (Song et al. 1997; Knemeyer et al. 1999). The ability of paclitaxel to distribute within tissue has also been studied indirectly using average cell survival in multicellular spheroids (Nicholson et al. 1997) and via flux through multilayered cell cultures (Nicholson et al. 1997; Tannock et al. 2002; Grantab et al. 2006), though without subsequent mathematical modelling to predict the tissue pharmacokinetics. Chapter 1 - Introduction 25 Both of the taxanes studied exert their anti-mitotic effects through binding and stabilization of microtubules leading to cell cycle arrest (Garcia et al. 1994). In the human colorectal carcinoma cell line used here, HCT-116, both drugs exhibit a similar IC 5 0 (Bissery 1995), though in general docetaxel exhibits a lower IC50 than paclitaxel. There are numerous differences between them which might affect their tissue penetration. Docetaxel has been reported to have a higher binding affinity to microtubules as well as faster accumulation and longer retention than paclitaxel (Kelland and Abel 1992; Lavelle et al. 1995) which might limit its ability to distribute within tissue. In addition, in vivo, both drugs are highly blood bound, paclitaxel 88-98% and docetaxel ~94% (Song et al. 1996; Sparreboom et al. 1998; Goodman et al. 2006). Assuming blood bound drug does not contribute to tissue build-up then the extent and speed at which binding occurs and the degree of reversibility for each drug could play an important role in determining the amount of free drug available for diffusion into tissue. Paclitaxel has the added factor of its formulation vehicle, Cremophor EL, which through micelle formation is thought to reduce the free drug available for diffusion but increase the duration of exposure (Gelderblom et al. 2002; ten Tije et al. 2003). Plan of work A - Initial work was carried out to determine conditions for drug exposure and end-point for measurement of drug distribution using the effect-based assay. Once it was determined that the assay could be implemented for both drugs, work was carried out to evaluate their ability to penetrate and distribute within tissue. B - Based on the significant differences between the two agents observed in part A we attempted to determine the distribution of drug effect on proliferation in tumour xenografts grown from the same cell line. Work was carried out where immunodetected BrdUrd labelled cells were mapped in relation to tumour vasculature. We then looked at the proliferation profile relative to tumour vasculature following drug treatment over a range of doses. C - We subsequently carried out radiolabelled based work to directly visualize drug profiles within MCCs and tumour xenografts, where the effect of vehicle Cremophor EL was also evalu-ated. The tissue distribution of 3H-paclitaxel and 3H-docetaxel was mapped in relation to tumour vasculature by combining a film-based dry-autoradiography technique with subsequent immu-nohistochemical detection of tumour vasculature. Superimposition and alignment of tumour vasculature and radiolabelled drug was made possible by capturing high-resolution images of entire tumour cryosections using a robotic tiling microscope. Following automated analysis of whole tumour cryosections, drug distribution was determined as a function of distance to the nearest visible blood vessel. While mapping radiolabelled taxane distributions relative to tumour vasculature will yield information about in vivo taxane distribution, it has the drawback of incorporating the contribution of 3H-labelled liver metabolites, which might diffuse more easily. Chapter 1 - Introduction 26 A second drawback of the technique is that drug penetration will likely appear to be greater than it actually is due to contributions from vessels that lie just outside of the tissue section, the occurrence of which will increase in likelihood with increasing distance from an observed vessel. These two factors mean that the in vivo based detection of drug penetration will likely over predict taxane tissue penetration. ANTI-CANCER DRUG SURVEY: EFFECT-BASED ASSAY In addition to the data presented on anthracyclines and taxanes, the tissue distribution of a range of other anti-cancer agents was surveyed using the effect-based assay. The objective was to evaluate the applicability of the effect-based assay for some of the more widely used anti-cancer agents. In addition to determining the usefulness and limitations of the assay, it was hoped that the survey would produce comparative estimates of the tissue distribution of some of the most commonly used anti-cancer drugs using a standardised assay. Drugs examined included representative agents from vinca alkaloids, antimetabolites, antibiotics, platinum agents, epiodophyllotoxins and alkylating agents. Vinca alkaloids The vinca alkaloids are a class of tubulin binding drugs of which several agents have been developed for clinical use. As such they present another opportunity to compare tissue distribu-tion of a group of related compounds possessing differing physicochemical properties. Due to their high level of cellular accumulation, up to 100-fold over extracellular levels (Ferguson and Cass 1985), and differing rates of cellular uptake it was hypothesised that they might exhibit different tissue penetration. Using the effect-based assay the tissue distribution of vincristine, vinblastine and vinorelbine was compared. Of the three, vincristine is the least lipophilic (Owellen et al. 1977) and has been shown to enter cells more slowly in comparison to vinblastine but is also retained longer (Ferguson and Cass 1985). Vinorelbine, which, is the most lipophilic of the three (Zhou et al. 1994), might benefit from an increase in access to the intracellular compart-ment if cellular sequestration or consumption do not play limiting roles in determining its tissue penetration. Antimetabolites The tissue distribution of three antimetabolites, gemcitabine, 5-fluorouracil (5-FU) and cytarabine was examined using the effect-based assay. All three drugs inhibit DNA and RNA function through their similarity with thymidine, in the case of 5-FU, and cytidine as with cytarabine and gemcitabine. While all three target cells during S-phase, the activity of gemcitabine, unlike cytarabine and 5-FU, is not solely confined to that portion of the cell cycle (Goodman et al. 2006). Additionally, cytarabine and 5-FU are commonly administered as continuous infusions and hence represent drugs that may not perform well using the short drug exposure required for the closed-off assay. Chapter 1 - Introduction 27 Others In addition to the drugs listed above a selection of other clinically used agents were tested using the assay including cisplatin, mitomycin C, bleomycin, etoposide, BCNA and thiotepa. Of these drugs mitomycin C, bleomycin and etoposide have all been reported to exhibit oxygen depend-ent toxicity and may exhibit gradients in effect which do not arise from limited diffusion and hence present test cases for the assay. Plan of work The work outline for this section was to follow a similar approach as was used for the anthracy-clines and taxanes. Initial work was carried out to determine conditions for drug exposure and end-point for measurement of drug distribution using the effect-based assay. Once it was determined that the assay could be implemented, work was carried out to evaluate each of the drugs. Chapter 2 - Materials & Methods 28 2. M A T E R I A L S & M E T H O D S 2.1 CELL CULTURE MONOLAYER CULTURE HCT-116 (human colorectal carcinoma) and SiHa (human cervix squamous carcinoma) cells were purchased from American Type Culture Collection. Cells were grown as monolayers using minimum essential medium (GIBCO/BRL, Burlington, ON, Canada) supplemented with 10% foetal bovine serumab (GIBCO/BRL) or 5% bovine growth serumc'd (HyClone, Logan, Utah) and passaged every 3-5 days for up to 4 months before being discarded. MULTILAYERED CELL CULTURE Standard tissue culture inserts (CM 12 mm, pore size 0.4 um, Millipore, Nepean, ON, Canada) were coated with 150 ul collagen (rat tail type I a b and calf skin type IIIc,d Sigma), dissolved in 0.01 M HCI and diluted 1:4 with 60% Ethanol to 0.75 mg/ml, and allowed to dry overnight. HCT-116 (7.5 x 10s cells) or SiHa (5x10s cells) in 0.5 ml growth medium, were then added to the inserts and incubated for 15-18 hours to allow the cells to attach. The cultures were then incubated for 2-3 days in custom-built growth vessels, FIGURE 2-1A, to form MCCs ~150 um in thickness. Each growth vessel contained a frame that held up to 8 inserts completely immersed in 130 ml of stirred growth medium (700 rpm, 25 mm stir bar) under continual gassing (5% C0 2 , balance air) at37°C. B FIGURE 2-1 Schematic of growth vessel and drug penetration apparatus used for M C C - b a s e d work. A | Configuration used during MCC growth, up to 8 MCCs could be held suspended in stirred growth medium. B | Apparatus used during drug exposure to one side, the membrane side of each MCC was clamped sideways against a polyacrylate block, with a layer of parafilm sandwiched in between to ensure a complete seal. MCC sides were defined as near (exposed) and far (closed-off). C | Apparatus used to conduct drug flux experiments and also for control experi-ments, where MCCs were exposed to drug from both sides but under different oxygenation and glucose/lactate conditions. 'BrdUrd work presented in chapter 3, banthracycline work presented in chapter 4 , ctaxane work of chapter 5 and dmiscellaneous work from chapter 6. Chapter 2 - Materials & Methods 29 2.2 BRDURD-PENETRATION IN M C C A N D XENOGRAFTS MICE A N D TUMORS Female nod/SCID mice were bred and maintained in our institutional animal facility in accor-dance with the Canadian Council on Animal Care guidelines. The experiments described here were approved by the Animal Care Committee of the University of British Columbia. Mice were allowed free access to standard laboratory rodent food and water. Human cervical carcinoma (SiHa) cells (50 ul of 20 x 106 cells/ml) were implanted sub-cutaneously into the sacral region of the mice. The mice were used at 12-16 weeks of age, and ranged in weight from 20-26 g. The weight of the excised tumours was 70 ± 30 mg (mean ± std. dev.). TUMOUR PENETRATION ASSAY Tumour bearing mice were administered 5-bromo-2-deoxyuridine, BrdUrd (Sigma Chemical, Oakville, ON, Canada), i.p. at 25, 50, 100, 400, 1000 and 2000 mg/kg 2 hours prior to sacrifice. BrdUrd was administered as a 30-mg/ml solution in saline. As a marker of blood perfusion, mice were intravenously administered 75 ul of 0.6mg/ml DiOC7(3) (Molecular Probes, Eugene, Oregon, USA) dissolved in 75% DMSO, 5 minutes prior to sacrifice. After excision, tumours were cooled to -20° C on an aluminium block, embedded in O.C.T. medium (Tissue-TEK, Torrance, CA, USA) and stored at -20° C until sectioning. An outline of the experimental plan for this section is shown in FIGURE 2-2. SiHa tumors xenografts were grown to -150 mm3 a* 2 hours 5 min ]<=> excise & freeze tumors immunostain S-phase cells (BrdUrd) image tissue & BrdUrd overlay and analyse BrdUrd relative vessels FIGURE 2-2 Workflow for BrdUrd tumour labelling experiments designed to examine the distribution of labelled S-phase cells in relation to vasculature. The fluorescent dye DiOC7(3), used to demark tumour blood vessels, was imaged before tissue immunostaining to avoid wash out and loss of signal. M C C - B A S E D BRDURD PENETRATION ASSAY MCCs were exposed to BrdUrd from one side using a polyacrylate jig, FIGURE 2-1B, designed such that the bottom of each insert was clamped against a flat block, with a layer of Parafilm (American National Can, Chicago, II, USA) sandwiched in between to ensure a complete seal. Each MCC was exposed to a 7.5 ml stirred reservoir kept under controlled gassing and tempera-ture. A silicone o-ring was used to seal the gap between the insert and the orifice in the reservoir. Chapter 2 - Materials & Methods 30 During the penetration assay growth medium was supplemented with excess glucose (10 g/1 d-glucose, Sigma) in attempt to sustain proliferation on the far side of the cultures (Mueller-Klieser et al. 1986). Once placed in the jig, MCCs were allowed to equilibrate for 45 minutes and then BrdUrd was added to the growth medium at concentrations of 1, 5, 25 and 100 uM. Control MCCs were exposed to 25 uM BrdUrd from both sides under otherwise similar conditions. After an exposure of 1 hour the cultures were removed, rinsed in PBS (phosphate buffered saline), embedded in O.C.T. and stored at -20° C until sectioning. An outline of the experimental plan is shown in FIGURE 2-3. Equilibrate 45 min. rinse & freeze ) immunostain S-phase cells (BrdUrd) image tissue & BrdUrd overlay and analyse BrdUrd relative to edges FIGURE 2-3 Workflow for BrdUrd MCC labelling experiment designed to examine the distribu-tion of labelled S-phase cells in relation to culture edge. Following growth to 150 um MCCs were transferred to penetration apparatus and equilibrated for 45 min prior to the 1-h exposure to BrdUrd from one side. Cultures were subsequently frozen, cryosectioned and immunostained to incorporated BrdUrd. PHARMACOKINETIC ASSAY Bromodeoxyuridine was administered i.p. at a dose of 100 mg/kg to six SCID mice. Blood samples of approximately 20 ul were taken from the tail vein of the mice at 5, 10, 20, 30, 60 and 120 minutes after administration. For the 5-30 minute time points the mice were divided into two groups and sampling was alternated between groups at each time point. All mice were sampled for the 60 and 120-minute time points. Following each time point, blood samples were weighed, diluted with 500 ul methanol, vortexed for 20 seconds and stored at 4° C. Upon completion of all time points, samples were centrifuged at 2500 g for 10 minutes and 450 ul of the supernatant was removed and evaporated to dryness using a centrifugal evaporator (1 hour at 40° C, Labconco, Kansas City, MO, USA). Samples were then reconstituted in 50 ul of 0.1 M ammonium acetate pH adjusted to 3.5 and then assayed via HPLC on the same day. HIGH PRESSURE LIQUID CHROMATOGRAPHY Chromatographic analysis was carried out with Waters equipment (Mississauga, ON, Canada), including a model 510 pump, model 712 WISP injector and model 996 photodiode array detector. A Symmetry C18 column (3.9 x 150 mm) was used for sample separation. With a mobile phase Chapter 2 - Materials & Methods 31 consisting of a mixture of 25% acetonitrile and 0.1 M ammonium acetate, pH 3.5 (0.15:0.85) flowing at 1.0 ml/min whereby BrdUrd eluted after 9.2 minutes. Samples of 30 jul were injected and absorbance detection was carried out at 280 nm. The assay linearity for BrdUrd in water was verified over the range of 0.2-200 uM and BrdUrd recovery from spiked blood was grater than 90% for 1-200 uM range. BrdUrd stability to within greater than 3% in the processed samples was confirmed over a 24 hour time period. B R D U R D IMMUNOHISTOCHEMISTRY MCC cryosections (10 um) were air dried for 24 hours and then fixed in a 1:1 mixture of acetone-methanol for 10 minutes at room temperature. Slides were immediately transferred to distilled water for 10 min and then treated with 2 M HCI at room temperature for 1 hour followed by neutralization for 5 min in 0.1 M sodium borate. Slides were then washed in distilled water and transferred to a PBS (phosphate buffered saline) bath. Subsequent steps were each followed by a 5 min wash in PBS. BrdUrd incorporated into DNA was detected using a 1:200 dilution of monoclonal mouse anti-BrdUrd (clone BU33, Sigma) followed by 1:100 dilution of anti-mouse peroxidase conjugate antibody (Sigma) and 1:10 dilution of metal enhanced DAB substrate (Pierce, Rockford, IL). Slides were then counterstained with haematoxylin, dehydrated and mounted using Permount (Fisher Scientific, Fair Lawn, NJ, USA). I M A G E ACQUISITION The imaging system consisted of a fluorescence microscope (Zeiss III RS, Oberkochen, Germany), a cooled, monochrome CCD video camera (model 4922, Cohu, San Diego, CA, USA), frame grabber (Scion, Frederick, Maryland, USA), a custom built motorized x-y stage and customized NIH-Image software (public domain program developed at the U.S. National Institutes of Health, available at http://rsb.info.nih.gov/nih-image/) running on a G4 Macintosh computer. The motorized stage allowed for tiling of adjacent microscope fields of view. Using the image acquisi-tion system, images of entire tumour sections were captured, typically 25-50 mm2 in size at a resolution of 1 pixel/um2. Two cryosections, typically 8 mm in length each, were imaged for each MCC, making a ~2 mm2 imaged tissue area. For tumour cryosections, images of DiOC7(3) fluorescence within the sections were obtained prior to BrdUrd immunostaining using a 450-480 nm excitation filter and a 525 nm long pass emission filter. Once immunostained and mounted, the slides were then revisited and bright field images of BrdUrd positive staining obtained. I M A G E ANALYSIS: TUMOURS Using the NIH-Image software application and user supplied algorithms, images of DiOC7(3) fluorescence and BrdUrd/tissue staining from each tumour section were overlaid and areas of necrosis and staining artefacts were removed. On the fluorescence image DiOC7(3) positive regions were then identified by selecting all pixels that were 20% above the maximum tissue background intensity. DiOC7(3) positive regions that were less than 20 um2 in size were consid-Chapter 2 - Materials & Methods 32 ered artefacts and removed from the analysis. On the bright field images, BrdUrd positive staining was identified by selecting pixels that were two and a half standard deviations above tissue background levels. Measuring the distance from each point in the tissue to the nearest DiOC7(3) positive pixel and noting if it were BrdUrd positive or negative then determined the relation between proliferation and distance to the nearest blood vessel. The data were tabulated so as to determine the fraction of BrdUrd positive pixels of the total number pixels found at each distance to a blood vessel. IMAGE ANALYSIS: M C C S BrdUrd positive staining was identified as above. For the MCCs that were sealed off from one side during exposure to BrdUrd, the edges of the MCCs were traced out and then the position of each point in the tissue was measured relative to the two edges (expressed as a fraction between 0 and 1) and tabulated along with whether or not the point was BrdUrd positive. The fraction of tissue positive for BrdUrd of the total tissue found at each position within the MCC, divided into 30 steps, was then calculated. For MCCs that were exposed to BrdUrd from both sides, the fraction of BrdUrd positive pixels at each distance away from the nearest edge was determined. BrdUrd-based drug penetration assay Grow cultures Drug exposure Detect effect U . U . U U Grow to 150 j/m 1 -3 days Equilibrate Drug exposure Rinse wells x3 Wash-out Rinse MCCs x2 45 min 1-2 h 15 sec 1h 15 sec 3 Change media 1&3h Incubate 1-3 days Label S-phase 100 /vM BrdUrd 4h FIGURE 2-4 Outline of the proposed effect-based assay for evaluation of drug distribution. Once cultures were grown to the desired thickness, they would be temporarily transferred to the penetration apparatus where drug exposure would occur. Following exposure and wash out, cultures were returned to the growth apparatus and incubated for a period of 1-3 days before labelling S-phase cells. 2.3 EFFECT-BASED MCC D R U G DISTRIBUTION ASSAY FIGURE 2-4 outlines the proposed effect-based assay for evaluation of drug distribution using a biological endpoint. Once grown, MCCs were exposed to drugs from one side using a polyacry-late jig, FIGURE 2-1B, designed such that the bottom of each insert was clamped sideways against a flat block, with a layer of Parafilm (American National Can, Chicago, II, USA) sandwiched in between to ensure a complete seal. Each MCC was exposed to a 2.5c,d or 7.5b-ml stirred reservoir ' B r d U r d work presented i n chapter 3, banthracycl ine work presented in chapter 4, ctaxane work of chapter 5 and ' 'miscellaneous work from chapter 6. Chapter 2 - Materials & Methods 33 kept under controlled gassing and temperature. A silicone o-ring was used to seal the gap between the insert and the orifice in the reservoir. Once placed in the jig, MCCs were allowed to equilibrate for 45 minutes and then drug was then added to the growth medium. After 1-2 h drug exposure, the reservoirs were rinsed twice with fresh medium. MCCs were then incubated for 1 hour prior to removal from the jig to allow initial drug wash out from the exposed-side only. For some experiments growth medium was changed halfway through the 1-hour wash-out periodc,d. MCCs were then removed, rinsed in fresh medium three times for 15 seconds and placed in a growth jar containing 2 untreated control MCCs, that were not subjected to closure. Separate control experiments were carried out to verify that closure of the MCCs did not, on its own, affect proliferation 1 or 3 days later. To reduce the build-up of washed out drug, medium was replaced at 1 and 3 hours following return to the growth jar and the incubation with control MCCs allowed for detection of potential drug effects to unexposed edges that were derived from drug wash out from the exposed sides. MCCs were left for 1-3 days from the time of beginning drug exposure to allow manifestation of drug effect, following which 100 uM BrdUrd (5-bromo-2-deoxyuridine, Sigma Chemical, Oakville, ON, Canada) was added to the growth medium and cultures incubated for 4 hours to label S-phase cells. MCCs were then removed, frozen in O.C.T. medium (Tissue-TEK, Torrance, CA, USA) and stored at -80° C until sectioning. MCC CONTROL ASSAY Control experiments were carried out using a dual-reservoir apparatus, FIGURE 2-1C, in which equal drug exposure was made to both sides of MCCs but under 20% oxygen, 1 g/L glucose in the first reservoir versus 0% oxygen, 0.1 g/L glucose and 1 mM lactate0 in the second reservoir. The low oxygen growth medium was gassed overnight at 0% 0 2 and then re-equilibrated for 1-hour upon insertion of the MCCs within the dual-reservoir apparatus. For these experiments, the medium used was glucose free DMEM GIBCO/BRL, Burlington, ON, Canada) supplemented with 10% foetal bovine serumab (GIBCO/BRL) or 5% bovine growth serumc'd, (HyClone, Logan, Utah). Sodium bicarbonate levels were set to produce pH 7.4 under 5% carbon dioxide. Glucose was then added to the oxygenated side to match physiological levels, 1 g/L. On the anoxic side, the glucose was derived from the growth serum. Drug exposure and wash out were carried out in the manner described in the previous section. BRDURD IMMUNOHISTOCHEMISTRY BrdURd immunostaining was carried out as described in section 2.2. "BrdUrd work presented in chapter 3, banthracycline work presented in chapter 4, ctaxane work of chapter 5 and ""miscellaneous work from chapter 6. Chapter 2 - Materials & Methods 34 IMAGE ACQUISITION1" The imaging system was the same as described in section 2.2. Images of MCC sections, 8 mm in length, were captured at a resolution of 1 pixel/um2. Two cryosections were imaged from each MCC. For MCCs immunostained for BrdUrd incorporation, bright field images were captured. For drug fluorescence visualization, MCCs were imaged using unmounted cryosections 1-hour after sectioning using a 510-555 nm excitation filter and a 575-640 nm emission filter. IMAGE ACQUISITION0 4"1 The imaging system consisted of a robotic fluorescence microscope (Zeiss Imager Z l , Oberko-chen, Germany), a cooled, monochrome CCD camera (Retiga 4000R, Qlmaging, Vancouver, BC, Canada), a motorized slide loader and x-y stage (Ludl Electronic Products, Hawthorne, NY, USA) and customized ImageJ software (public domain program developed at the U.S. National Institutes of Health, available at http:/ / rsb.info.nih.gov/ij/) running on a G5 Macintosh com-puter (Apple, Cupertino, CA, USA). The system allowed tiling of adjacent microscope fields of view. Using this system, images of entire tumour cryosections up to 1 cm2 and MCC sections, 0.8 cm in length, were captured at a resolution of 0.75 um/pixel. IMAGE ANALYSIS Using the NIH-Image software application and user supplied algorithms, digital images of BrdUrd staining and drug fluorescence within MCC cryosections were analyzed. Pixels making up the cryosection were first sorted based on their distance relative to either edge of the tissue. For BrdUrd images, the fraction of positively stained pixels at each position relative to the two edges was then calculated. Pixels two and a half standard deviations above tissue background levels were classified as BrdUrd positive. For fluorescence images, the average intensity of pixels from each group relative to the two edges was calculated. Detection of tissue edges was facili-tated by capturing high-contrast, bright-field images of the MCCs (prior to staining and mounting the slides in the case of the BrdUrd sections). 2.4 ANTHRACYCLINE FLUX EXPERIMENTS FLUX ASSAY Flux experiments were performed using the dual reservoir diffusion apparatus shown in FIGURE 2-1C. Prior to the experiment cultures were selected and visually assessed for uniformity. The selected cultures were then placed in the apparatus with 7 ml medium per reservoir. Each reservoir was individually gassed, 10 ml/min, and maintained at 37°C via forced air heating. "BrdUrd work presented in chapter 3, banthracycline work presented in chapter 4, laxane work of chapter 5 and ""miscellaneous work from chapter 6. Chapter 2 - Materials & Methods 35 Each reservoir was stirred using a magnetic stir bar (15 mm x 4 mm) at 450 r.p.m.. Following a 45 min equilibration 100 uM drug was added and growth medium reservoirs were sampled, 100 ul, periodically over a 3-hour period. Following experiments, MCCs were frozen and sectioned to determine culture thickness. H IGH PRESSURE LIQUID CHROMATOGRAPHY HPLC analysis of samples taken from the flux apparatus reservoirs was generally begun midway through drug flux experiments. Analysis used a Waters HPLC system (Mississauga, ON) which included a model 510 pump, model 712 WISP Injector, a model 996 Photo-Diode Array (PDA) Detector and a 474 Scanning Fluorescence Detector. All eluents were made using HPLC grade products and distilled and de-ionised water was used for all dilutions. Before running sample analysis, HPLC eluents were degassed via vacuum filtration (0.5 um FHLP filter, Millipore, Mississauga, ON). Sample Preparation Reservoir medium samples were prepared for HPLC analysis immediately following their acquisition. Samples were de-proteinated by adding an aliquot of a 0.1-M citrate buffer pH 3.0 (volume equal to 10% of sample volume) followed by methanol (5x sample volume). After shaking, the protein precipitate was pelleted by centrifugation for 10 minutes at 7 x 103 r.p.m. (radial distance of ~5 cm). Samples were assayed via HPLC on the same day of each flux experi-ment. Analysis A Symmetry C18 column (3.9 x 150 mm, Waters, Mississauga, ON) was used for sample separa-tion. Samples of 30 ul were injected using a mobile phase of 90% acetonitrile/buffered H 2 0 mixture (0.3:0.7) flowing at 1 ml/min, whereby drugs eluted after ~4 minutes. The buffered H 2 0 eluent consisted of 0.1 M ammonium acetate, pH adjusted to 4 with HC1. Fluorescence detection was carried out using the 474 Scanning Fluorescence Detector using an excitation wavelength of 480nm and an emission wavelength of 560 nm (bandwidth 40 nm). For mitoxantrone, absorbance at 640 nm was used for detection. The assay linearity for the anthracyclines in spiked growth media was verified over the range of 0.01-100 uM for DOX, EPI and DAU and 0.025-100 uM for MIT. Stability in the processed samples to within 5% was confirmed over a 24 hour time period. 2.5 3 h - T A X A N E A N D TAXANE-TUMOUR XENOGRAFT WORK REAGENTS Stock solutions were prepared of paclitaxel (Sigma-Aldrich, Oakville, ON, CA) in Tween 80 (Sigma-Aldrich) at 12mg/ml and Cremophor EL (Sigma-Aldrich) at 6mg/ml and docetaxel (Sigma-Aldrich) in Tween 80 at 40mg/ml. For tritium based experiments, 3H-paclitaxel (2-benzoyl ring-3H, ART-1195, American Radiolabelled Chemicals inc. St Louis, MO, USA) in EtOAc Chapter 2 - Materials & Methods 36 and 3H-docetaxel (2-benzoyl ring-3H, ART-1267, American Radiolabelled Chemicals inc.) in EtOH were centrifugally evaporated at 30° C and mixed with cold drug stock at a ratio of 60 uCi per mg cold drug immediately prior to diluting for injection. For paclitaxel in Cremophor EL, stock was then diluted 1:1 in saline before injection for a final concentration of 3 mg/ml. For paclitaxel in Tween 80, stock was diluted 1:1 in EtOH and then 1:1 again in saline (Sparreboom et al. 1996) for a final concentration of 3 mg/ml. For docetaxel, stock was diluted 1:3 in 13% EtOH and then 1:25 in saline, final concentration 1.38 mg/ml. excise & freeze tumours overlay & analyse FIGURE 2-5 Radiolabelled taxane tumour mapping protocol. HCT-116 tumours xenografts grown in NOD/SCID mice were treated with 3H-paclitaxel and docetaxel i.v. and at 2 or 8 hours following treatment tumours were excised, frozen and cryosectioned. Following imaging of the blood flow marker DiOC7(3), tumour sections were clamped against tritium sensitive film and left for 80 days. Film was then developed and slides immunostained for the endothelial cell marker CD31. Imaging of entire tumour cryosections allowed subsequent alignment of film and tissue images. M ICE A N D TUMOURS Female NOD/SCID mice were bred and maintained in our institutional animal facility in accor-dance with the Canadian Council on Animal Care guidelines. Experiments were approved by the Animal Care Committee of the University of British Columbia. Mice were allowed free access to standard laboratory rodent food and water and were used between 8-11 weeks of age, weighing between 20 to 28 g. For radiolabelled-based work HCT-116 cells (50 ul of 1.6 x 108 cells/ml) were implanted sub-cutaneously into the left and right sacral regions, with 2 mice for a total of four tumours used for each data point. An outline of the experimental plan for radiolabelled work is shown in FIGURE 2-5. For BrdUrd-based work, outlined in FIGURE 2-6, only one tumour was implanted per mouse, with five mice per data point. Mice were administered drug when one or both tumours reached 100-150 mm3 as calculated from calliper measurement of three orthogonal diameters (a,b,c) using the formula volume = rc/6(abc). The mean weight (± SD) of the excised tumours was 60 ± 24 mg. For radiolabelled-based work, drugs were administered i.v.. For BrdUrd-based work, paclitaxel in Cremophor El and docetaxel in Tween 80 were administered i.p. and mice were administered 1500 mg/kg BrdUrd and 60 mg/kg pimonidazole (Hy-HCT-116 tumours e-r>' were grown to <uf$^ -150 mm* — ^ rj. JJ f - f A | 2,8 hours 5 min |1Z^ ) ^ > image blood flow (DIOC7(3» dry-autoradiography 80 days immunostain & image vessels (CD31) develop film Chapter 2 - Materials & Methods 37 poxyprobe-1 Kit, Chemicon International Inc., Temecula, CA, USA) 2 hours prior to sacrifice, as previously described (Huxham et al. 2004). HCT-116 tumours were grown to -150 m m 3 24-144 hrs j l — \ e 2 hrs 5 min tumours immunostain & image vessels, hypoxia (CD31,pimo) immunostain & image S-phase cells (BrdUrd) FIGURE 2-6 Protocol for mapping the effect of taxanes on S-phase cell distribution tumours. HCT-116 tumour xenografts grown in NOD/SCID mice were treated with 3H-paclitaxel and docetaxel i.p. and at 24-144 hours following treatment tumours were excised, frozen and cryosec-tioned. Following imaging of the blood flow marker DiOC7(3), tumour sections were immunostained for the endothelial cell marker CD31 and hypoxia, via detection of pimonidazole adducts. Slides were then immunostained for S-phase cells via detection of incorporated BrdUrd. Imaging of entire tumour cryosections allowed subsequent alignment of the images. RADIOLABELLED TAXANES IN MCCS MCC drug penetration was performed using the apparatus shown in FIGURE 2-1B as described above but with cultures frozen immediately following the 1 or 2-hour drug exposures with no rinsing. An outline of the experimental plan is shown in FIGURE 2-7. 10 p thick tumour xenograft and MCCs cryosections were placed in triplicate on glass slides, allowed to dry overnight, and then clamped against tritium sensitive film (Hyperfilm-3H, GE Healthcare, England) under GBX filter safelight conditions (Kodak, Rochester, NY, USA) and stored in darkness for a period of 80 days. Film was developed for 1 minute using a 1:1 dilution of D-19 developer (Kodak). Calibration of film was performed using a tritium standard slide, range 0-489 uCi/g (American Radiolabelled Chemicals inc., St Louis, MO, USA). Freeze & Section •II Overlay & Analyse FIGURE 2-7 Radiolabelled taxane MCC mapping protocol. Following growth to 150 um cultures were transferred to penetration apparatus and equilibrated for 45 min prior to the l&2-h expo-sures to 3H-paclitaxel and docetaxel from one side. Cultures were subsequently frozen and cryosectioned. Slides were then clamped against tritium sensitive film and left for 80 days after which film was then developed. Imaging of entire cryosections allowed subsequent alignment of film and tissue images. Chapter 2 - Materials & Methods 38 IMMUNOHISTOCHEMISTRY Cryosections were fixed in acetone-methanol for 10 minutes at room temperature. Vasculature was stained using an antibody to PECAM/CD31 (BD PharMingen, San Diego, CA, USA) and fluorescent Alexa 546 anti-rat secondary (Invitrogen, Burlington, ON, CA). Hypoxia was detected via bound pimonidazole adducts using Hypoxyprobe Mabl (Chemicon) and an Alexa 488 anti-mouse secondary. Slides were subsequently stained for incorporated BrdUrd following the method described in section 2.2 (proceeding directly to the first distilled water wash). IMAGE ANALYSIS Using the NIH-Image software application and user supplied algorithms, digital images of 3 H -taxane distributions were analyzed. For tumours, digital images of tritium sensitive film and CD31 immunostained tumour cryosections were superimposed and aligned. Then each pixel making up the film image was sorted based on its distance relative to CD31 positive vasculature. Finally, the fraction of pixels from the film that were positive was tabulated as a function of distance from vasculature. Data were scaled using results for fraction positive pixels versus radioactivity (uCi/g) using a tritium standard slide. For MCCs, digital images of tritium sensitive film were aligned with bright field images of unstained tissue cryosections. Pixels making up the tissue were then sorted based on distance from each edge of the section. For BrdUrd-based work, analysis was carried out as previously described (Huxham et al. 2004). Chapter 3 - Bromodeoxyuridine 39 3. BROMODEOXYURIDINE 3.1 SUMMARY Bromodeoxyuridine (BrdUrd) is used extensively to measure the fraction of proliferating cells in tumours. Unlike endogenous markers of proliferation, such as proliferating cell nuclear antigen (PCNA) and Ki-67, BrdUrd is exogenously administered and reaches the tumour via vasculature where it must then diffuse throughout the tissue in order to label S-phase cells. In this study we examine the dose dependence of BrdUrd on the tumour distribution of labelled cells in histologi-cal sections. Analysis of the distribution of labelled cells in SiHa tumour xenografts showed that a dose between 400 and 1000 mg/kg was required to label S-phase cells 150 um from blood vessels, approaching the border of necrosis. Lower doses resulted in only the cells close to blood vessels being labelled. Cells residing 150 um from blood vessels were found to label at half the level of those situated in close proximity to the tumour vasculature. Results were compared with the penetration of BrdUrd seen in vitro using multilayered cell culture (MCC). Using MCC, an exposure of 100 uM BrdUrd for 1 hour was required for labelling of S-phase cells 150 um into the tissue, while cells adjacent to the edge of the tissue could be adequately labelled with just 5 uM BrdUrd for 1 hour. The plasma AUC for a 100-mg/kg BrdUrd dose in mice was found to be -30 uMxh. 3.2 RESULTS BRDURD PENETRATION INTO TUMOR XENOGRAFTS The distribution of BrdUrd labelled S-phase cells around tumour micro-vessels was assessed via simultaneous detection of BrdUrd-DNA adducts and DiOC7(3) stained blood vessels in tumour cryosections. BrdUrd was injected i.p. at four doses 100, 400,1000, 2000 mg/kg to multiple mice 2 hours prior to sacrifice; single mice were dosed at 25 and 50 mg/kg. Images of whole tumour cryosections were captured for both markers and then overlaid. Panels A-D of FIGURE 3-1 show sample areas from images of the 100-2000 mg/kg dosed mice. Results of analyses to determine the distribution of BrdUrd labelling relative to tumour vasculature from images of entire cryosec-tions are shown in panels E-H. At all doses a trend of decreasing BrdUrd staining with distance away from vasculature was seen. As dose increased, an increase was seen in the intensity of staining and the number of cells exhibiting staining. This was interpreted as being due to the higher plasma BrdUrd concentrations and the longer period over which labelling could be expected to occur at the higher doses. Chapter 3 - Bromodeoxyuridine 40 3* •Ifv 50 100 150 0 50 100 150 0 50 100 150 0 Distance to nearest visible blood vessel (um) 150 FIGURE 3-1 SiHa tumour cryosections showing S-phase labelled cells (non-thresholded, greyscale) relative to vasculature (thresholded, magenta) as a function of BrdUrd dose. A1100, B|400, C11000 and D | 2000 mg/kg BrdUrd. Fluorescence images of DiOC7(3) demarcated vasculature were thresholded and overlaid on subsequent bright-field images of cryosections immunostained to show BrdUrd labelled S-phase cells counterstained with haematoxylin (non-thresholded). Inlays show magnified regions. Scale bars 150 um. Line graphs show the distribu-tion of BrdUrd labelling (fraction tissue identified as being BrdUrd positive, see section 2.2) as a function of distance to nearest visible blood vessel at the four doses, E | 100, F1400, G11000 and H12000 mg/kg BrdUrd. Each line represents data from an individual tumour. Chapter 3 - Bromodeoxyuridine 41 An increase was also seen in the penetration of BrdUrd away from the vasculature with increas-ing dose. From the distribution plots, the depth at which proliferation appeared to drop to half-maximum was determined as 60 ± 10 um (n=2), 120 ± 10 um (n=2), 135 ± 25 um (n=4) and 110 ± 20 um (n=4) (mean ± max. deviation) for the 100, 400, 1000 and 2000-mg/kg doses respectively; the 400 and 1000 mg/kg groups each contained a tumour that did not follow the general trend and was excluded from the half-maximum calculations. The half-maximum values from the highest three doses were within experimental error of each other, however peak staining levels were significantly higher for the 1000 and 2000-mg/kg doses. Single mice dosed with 25 and 50 mg/kg continued the trend towards reduced staining and penetration (data not shown). Results indicated that a dose between 400 and 1000 mg/kg was required for maximal immunohistochemical staining of S-phase cells located far from blood vessels in this tumour model. BRDURD PHARMACOKINETICS IN SCID MICE The plasma concentration of BrdUrd in SCID mice was measured at several times after i.p. administration of 100 mg/kg. BrdUrd blood concentration versus time is shown in FIGURE 3-2. The rate of disappearance was found to be exponential over the first 2 hours with a half-life of 19 ± 1 (SE) minutes. The peak level was determined as 58 ± 12 (SE) uM and the area under the curve for the first 2 hours was ~30 uMxh. A well defined absorption phase was not observed from these data. 100 r. • 1 • 1 q Time post injection (hours) FIGURE 3-2 BrdUrd plasma kinetics in SCID mice after intraperitonal administration of 100 mg/kg i.p.. Plasma levels follow an exponential decay over the first 2 hours. The half-life was determined to be 19 ± 1 (SE) minutes, peak plasma levels were 58 ± 12 (SE) uM and the area under the curve for the first 2 hours was -30 uMxh. Points show mean ± SE (n= 3-6). Chapter 3 - Bromodeoxyuridine 42 BRDURD PENETRATION INTO MCC The penetration of BrdUrd into MCC was assessed via immunohistochemical detection of BrdUrd-DNA adducts in cultures exposed to BrdUrd at varying concentrations. Exposure to BrdUrd was made from one side of the cultures in order to observe penetration of BrdUrd through the entire width of the culture and enable comparison of BrdUrd staining on either edge, where cells were expected to display similar rates of proliferation. FIGURE 3-3 shows sample areas of MCCs exposed to 1, 5, 25 and 100 uM BrdUrd for 1 hour an additional MCC was exposed to 25 uM from both sides. The images show that 100 uM BrdUrd is required to label S-phase cells on the far edge of a 150-um thick culture. Results from analysis of the relation between BrdUrd labelling and position within the cultures, performed on images of the entire cryosections, are also shown. Data show that at 100 uM BrdUrd, labelled cells are uniformly distributed within the culture and match the labelling seen after exposure to 25 uM BrdUrd from both sides. CD > "to o Q. "O i— p CD => CO CO c o o CO I, p \ • 1 pM 25 (pM : . •• Relative position within M C C 150 I direction of 1150 £ drug diffusion FIGURE 3-3 Greyscale images of 3-day-old MCC showing the distribution of S-phase labelled cells as a function of BrdUrd concentration. 1 h exposures to 1-100 uM BrdUrd were made from the top-side of the cultures and additional cultures were exposed to 25 uM BrdUrd from both sides. S-phase cells were detected via immunohistochemical staining of BrdUrd-DNA adducts. Cryosections were counterstained with haematoxylin. Scale bars 150 um. Line graphs show the distribution of BrdUrd labelling for corresponding cultures. Data are expressed as a function of position relative to each edge of the cultures, 0 indicating cells adjacent the top (exposed) edge and 1 indicating the cells adjacent the bottom (far) edge of the cultures. Each line represents data from an individual cryosection. Chapter 3 - Bromodeoxyuridine 43 In order to observe the effect of exposure time on the penetration of BrdUrd at low concentra-tions, additional cultures were exposed to BrdUrd for 24 hours. In this case, the cultures were exposed to BrdUrd from both sides in order to avoid loss of proliferation on the far edge due to closing off the culture over the extended period. MCCs were exposed to 1 uM BrdUrd for 2 hours, 1 uM BrdUrd for 24 hours and 1 uM BrdUrd for 24 hours followed by 200 JUM BrdUrd for 1-2 hours to detect S-phase cells not labelled by the 24-hour exposure at the low concentration. During the 24-hour exposure BrdUrd concentrations were monitored by HPLC and maintained at 1 ± 0.5 uM (BrdUrd fell to ~0.4 uM after 12 hours at which point it was supplemented to 1.5 iuM). FIGURE 3-4 shows sample areas of the immunostained MCC cryosections. Due to the additional day of growth the MCCs are thicker than those from the 1-hour exposures and a loss of proliferation occurs towards their centres. Results indicate a marked increase in BrdUrd penetration after exposure to 1 uM BrdUrd for 24 hours versus 2 hours. However, labelling in the central region of the MCCs was seen to increase after exposure to 200 uM BrdUrd for 2 hours suggesting inadequate penetration of 1 uM BrdUrd even after 24 hours. Analysis of the distribu-tion of BrdUrd labelling as a function of distance to the nearest edge of the tissue is shown in the lower panel of FIGURE 3-4. A marked increase in labelling is seen throughout the MCCs after exposure to BrdUrd for 24 hours in comparison to the 2-hour data and an additional increase in labelling at a depth of 150 um is seen upon subsequent exposure to 200 uM BrdUrd. Chapter 3 - Bromodeoxyuridine 44 Distance to nearest edge (um) FIGURE 3-4 Greyscale images of 4-day-old MCC showing distribution of BrdUrd labelled S-phase cells after extended exposure from both sides under various conditions: 1 uM BrdUrd for 2 hours, 1 uM BrdUrd for 24 hours, 1 uM BrdUrd for 24 hours followed by 200 uM BrdUrd for 1 hour and 1 uM BrdUrd for 24 hours followed by 200 uM BrdUrd for 2 hours. Cryosections were counterstained with haematoxylin. Scale bars 150 um. Lower panel shows analysis of the distri-bution of BrdUrd labelling as a function of distance to nearest edge of the cultures. Error bars show standard deviation. Chapter 3 - Bromodeoxyuridine 45 3.3 DISCUSSION The distribution of BrdUrd labelled S-phase cells following BrdUrd administered by intra-peritoneal injection over a range of doses was evaluated using a human tumour xenograft grown in mice. Results indicated that a dose of between 400 and 1000 mg/kg BrdUrd was required to adequately label S-phase cells located distal to vasculature. We compared the results from tumour cryosections with data from multilayered cell culture (MCC) experiments using a new method which circumvented the existence of gradients in proliferation seen in vivo. MCCs exposed to increasing concentrations of BrdUrd for 1 hour showed a trend in tissue distribution versus exposure that was consistent with that seen from the tumour data. MCC results indicated that a BrdUrd exposure of approximately 100 uM for 1 hour was required to label S-phase cells residing 150 pm from the site of exposure. It appeared that at the highest two doses, both in the tumour and MCC studies, the maximal rate of BrdUrd incorporation into DNA was being achieved throughout the tissue for the given proliferation status. At these doses the actual distribution of free BrdUrd was unknown and a gradient may still have existed within the tissue. A relationship between increasing dose of BrdUrd and increasing labelling has previously been shown (Peschke et al. 1999) and has generally been attributed to increased staining intensity and increased number of cells entering S-phase during the longer exposure to relevant BrdUrd levels. This study shows that there is an additional tissue penetration component, characterized by an increase in the labelling of cells far from blood vessels with increasing BrdUrd dose. The finding that a dose of BrdUrd between 400 and 1000 mg/kg was required to label S-phase cells furthest from tumour vasculature in cryosections was initially surprising since most flow cytometry studies are performed with doses of 100 mg/kg or lower. Flow cytometry can have a higher signal to noise ratio compared to immunohistochemical studies as is illustrated by data from the literature for SiHa xenografts showing a labelling index of approximately 25% for a 90 mg/kg dose of BrdUrd (Sham and Durand 1999), which is comparable to results shown here for the 1000-mg/kg dose. However, a BrdUrd-unlabeled population of cells containing S-phase levels of DNA is commonly observed in flow cytometric studies (Begg et al. 1985; Wilson et al. 1985; Rew et al. 1992) examining tumour cell proliferation. The presence of the unlabelled cells has generally been interpreted as indicating non-cycling cells with intermediate levels of DNA rather than unlabeled and proliferating cells. BrdUrd labelling of tumour tissue in vitro has been shown in some instances to increase the labelling index detected via flow cytometry (Wilson et al. 1985). Also, a study by Rodrigez using iododeoxyuridine in the HCT-116 xenograft, which exhibits similar vascular density to SiHa xenografts, found that after continuous infusion for 5 days at 100 mg/kg/day, immunohistochemical staining clearly showed incomplete tissue staining while flow cytometric data indicated close to a 90% labelling index (Rodriguez et al. 1994). Human studies involving BrdUrd as a diagnostic tool have typically used much lower doses of the agent, generally 100-250 mg/patient, (approx. 1-3 mg/kg) (Wilson et al. 1985; Tsang Chapter 3 - Bromodeoxyuridine 46 et al. 2000; Zackrisson et al. 2002) compared with the 400-1000 m g / k g needed in our studies in mice. Based on the findings presented here it seems likely that these low doses may underesti-mate the fraction and distribution of S-phase cells in solid tumours. Using M C C s , a 1-hour exposure time to BrdUrd was chosen to match the exposure for typical bolus injections used in mice. Results indicate that BrdUrd concentration must be kept above 100 u M for 1 hour in order to adequately label S-phase cells 150 um from the site of exposure. The reason for poor penetration at lower concentrations is interpreted as being due to consumption of BrdUrd by the tissue. Extended exposure to 1 uM BrdUrd , FIGURE 3-4, improved its tissue penetration relative to the 1-hour data, but was still unable to produce maximal labelling in the cells furthest removed from the site of exposure. The M C C data showed that BrdUrd penetration, and not solely the reduction in proliferation away from blood vessels, could be involved in determining the labelling pattern seen i n the rumour xenografts. The clinical testing of BrdUrd as a radiation sensitizer in the 1980-90's generally utilized continu-ous infusion over several days, wi th doses of ~1000 m g / m 2 / d a y (equivalent to approx. 25 mg /kg /day ) . A t these dose levels, BrdUrd reached steady state plasma concentration of ~1 u M (Kinsella et al. 1984a). In the M C C system a 24-hour exposure to 1 u M BrdUrd did not produce maximal labelling of proliferating cells distant from the site of exposure, indicating sub-optimal exposure due either to limited tissue penetration or cellular uptake of drug. It is possible that an extended exposure to 1 u M BrdUrd may be insufficient to fully sensitize proliferating cells located distant from vasculature. M C C s have in the past been used to evaluate drug penetration via measurement of flux through the cultures (Hicks et al. 1998; Kyle and Minchinton 1999; Tunggal et al. 2000; Hicks et al. 2001; Tannock et al. 2002). The data presented here represents a new application of M C C s in the evaluation of drug penetration via immunodetection of drug-cell interactions within histological sections. Evaluating a drug's distribution via its effect is usually confounded by intrinsic changes in cells wi th depth into tissue (e.g. reduced proliferation, drug sensitivity, hypoxia, necrosis) (Sutherland 1988; Olive and Durand 1994). In this study, this problem was circumvented by temporarily closing off one side of the culture during exposure from the other side and then examining the cell layers adjacent to each surface. The purpose of this approach was to attempt to allow comparison of BrdUrd labelling in two cell populations with similar rates of proliferation. This approach could be used to evaluate the distribution of anticancer agents based on their effect on cells located at different depths in tissue using immunohistochemical assays for proliferation or cell death. Chapter 4 - Anthracyclines 47 4. ANTHRACYCLINES 4.1 S U M M A R Y The failure of many anticancer drugs to control growth of solid cancers may stem in part from inadequate delivery to tumour regions distant from vasculature. While the identification of new anticancer drug targets has led to the development of many new drug candidates, there is currently a lack of methodology for studying the role played by drug penetration in determining their activity. In this chapter the feasibility of using multilayered cell culture (MCC) in conjunc-tion with an effect-based assay to evaluate the ability of drugs to penetrate and distribute within tissue is explored. MCCs were grown to ~150 um in thickness and then exposed to anthracyclines with one side temporarily closed off. Penetration was then assessed 1 to 3 days later by looking at the differential effect on the exposed and far sides of the cultures of the drugs via BrdUrd-based detection of S-phase cells. We chose to study the relative distribution of four anthracyclines, doxorubicin (DOX), epirubicin (EPI), daunorubicin (DAU) and the related compound mitoxan-trone (MIT). Initial work was conducted to evaluate the effect of a 1-4 hour closure of the MCCs on prolifera-tion status within the cultures. Following this, a pilot study was carried out with DOX alone to determine appropriate concentrations and waiting periods to produce detectable effects using the assay. Upon optimisation of the assay the tissue distribution of the selected anthracycline analogues was assessed. At clinically relevant exposures, none of the agents was able to affect cells on the far side of the culture at levels approaching that seen on the near (exposed) side. DOX and EPI exhibited ~10-fold decreases in the drug exposure seen by the cells on the far side relative to those on the near side of the cultures, while for DAU and MIT ~30-fold and >30-fold decreases were observed respectively. Results from the effect based assay were then compared with direct imaging of drug-derived fluorescence as well as ranking of drugs based on their flux through MCCs. Results from drug-derived fluorescence of DOX, EPI and DAU were found to be consistent with effect-based data. Flux experiments, which measured the relative flux of the four anthracyclines through MCCs, were found to rank the anthracyclines in order of their observed distributions. Additional work was carried out to examine the effect of inhibition of drug seques-tration in acidic lysosomes/endosomes. Results showed that while inhibition of sequestration reduced net accumulation of the drug it did not appear to increase tissue penetration. 4.2 RESULTS P R E L I M I N A R Y F INDINGS: F L U X - B A S E D E V A L U A T I O N O F A N T H R A C Y C L I N E DIFFUSION T H R O U G H M C C S Prior to conducting BrdUrd-based MCC experiments, the flux through MCCs of the four anthra-cyclines was measured. This initial work was carried out using MCC grown with SiHa (human squamous carcinoma) cells rather than HCT-116 cells, which were chosen for the subsequent Chapter 4 - Anthracyclines 48 BrdUrd-based work due to their more consistent growth for both MCC and tumour based work. FIGURE 4-1 summarizes results from a series of experiments in which anthracycline flux through MCC was measured. From the data, doxorubicin, followed by epirubicin, were seen to exhibit a greater rate of flux through tissue than daunorubicin and mitoxantrone. In the graphs the rate of appearance of drug in the receiving reservoir is shown as percent of the driving concentration in the donating reservoir. When the lysosomotropic base NhLCl was added to both reservoirs, under which the process of acidic sequestration of the drugs is meant to be inhibited, the flux of all four drugs was seen to increase dramatically. This increase in drug flux suggests that cellular accumulation via lysosomic sequestration could be one factor that limits anthracycline tissue penetration. However this work was carried out at relatively high concentrations, 100 uM drug (discussed below), and work shown later in this chapter at 10 uM drug did not confirm these results. Net metabolites refers to fluorescent peaks that appeared in HPLC chromatograms that were not present at time zero. For interest sake, the area of these peaks were summed and normalised to the peak area of the driving concentration in the same manner of the parent drug. Separate experiments using inserts without cells found all four drugs reached ~18% of the driving concentration after 180 min (DOX 19.7±1%, EPI 18±1%, DAU 17.5±1%, MIT 18±1%). Drug loss through hydrolysis was determined in separate experiments to be ~5% over the 180 minute period. Concentrations following incubation of 100 uM drug in the growth medium at 37° C for 180 min were DOX 97±5 u.M, EPI 94±5 uM, DAU 94±5 uM, MIT 100±5 uM. Loss in the donating reservoirs that could be attributed tb flux into the cultures was below the reproducibility of the assay. Chapter 4 - Anthracyclines 49 Flux assay: measure appearance of drug in receiving reservoir normal conditions + 50 mM NHGI 4 D) •a 0.75 ' dox • parent % donating reservoir cone, epi . * * ~ dau —•— mit % donating reservoir cone, % donating reservoir cone, 0.15 6 O) c c o 1 2 0.1 II S? I 0.05 o 0 0 60 120 180 0 60 120 180 Time (min.) Time (min.) FIGURE 4-1 Flux data showing the rate of appearance of anthracyclines and their metabolites in the receiving reservoir following diffusion through SiHa MCCs. All experiments were carried out using 100 pM drug in the donating reservoir, with results plotted as percent of the starting concentration. Data shown are averages from several experiments. Points show mean ± SE (n=5-8). (DOX: MCC thickness 180 ± 15 um, n=7; EPI: 165 ± 15 pm, n=8; DAU: 165 ± 15 pm, n=7; MIT: 175 ± 5 pm). Images of drug-derived fluorescence in cryosections taken from the cultures used for the flux assay are shown in FIGURE 4-2. Results indicate that the rate of flux through the cultures ranks inversely to the amount of fluorescence seen in the cryosections. Having the culture open to stirred reservoirs on both sides will in theory reduce build-up within the culture because the far side should follow the concentration seen in the receiving reservoir. For example at the end of the doxorubicin flux assay the receiving reservoir is only at ~0.25% of the 100-pM driving concentra-tion (0.25 pM). Under conditions of 50 mM NH4C1, anthracycline accumulation is seen to be reduced, which is consistent with the increased flux measured through the tissue shown in FIGURE 4-1 and the overall expectation of reduced sequestration of the drugs. However, the cultures also appear to be less intact at the end of the experiment as seen from examination of the haematoxylin images shown in FIGURE 4-2. Hence, a compromised barrier to diffusion may be responsible for the increased flux, this second explanation is consistent with data conducted at 10 pM drug pre-sented later in this chapter (see FIGURE 4-9). The far edge of the daunorubicin culture under NH4C1 does not appear to follow the receiving reservoir concentration, which reaches only ~0.2 uM at the end of the experiment versus driving concentration on the other side of 100 pM. Chapter 4 - Anthracyclines 50 F I G U R E 4-2 Anthracycline fluorescence in 10 um cryosections taken from SiHa MCCs exposed to 100 uM drug for 3 hours under conditions used for flux experiments. Colour images on left show haematoxylin staining of the tissue carried out after fluorescence imaging. Traces on the far right show fluorescence intensify as a function of position within each MCC. EFFECT B A S E D E V A L U A T I O N O F A N T H R A C Y C L I N E P E N E T R A T I O N IN MCCs The effect of the anthracyclines on S-phase labelling on either edge of MCCs was initially exam-ined over a range of concentrations (0 to 100 uM), drug exposures (1-3 hours), incubation periods following exposures (1-3 days) and growth conditions before a limited range of each parameter was selected for the study. Drug exposures were made to one side of the MCCs using the apparatus shown in FIGURE 2 - 1 B ; the two sides of the MCCs were defined as the near side (exposed) and far side (closed-off). Exposure from one side enabled a comparison of drug effect on cells that displayed similar rates of proliferation that were either directly exposed to drug in medium or separated from the drug by the thickness of the culture itself. Following drug expo-Chapter 4 - Anthracyclines 51 sures, cultures were returned to their growth vessels and incubated in fresh medium to allow manifestation of drug effect within the tissue. Results from preliminary experiments to show proof of principle for the assay are shown in FIGURE 4-3. untreated 5 uM dox 10 p.M dox ••• • . . . . . . . *= 100 uM dox position in M C C 150 IJ. direct ion of drug diffusion FIGURE 4-3 Proof of principle of BrdUrd-based drug penetration assay. HCT-116 MCC grown to -175 um thick were exposed to increasing concentrations of doxorubicin for 1-hour and then incubated for 1-day prior to detecting S-phase cell distribution within the cultures. Results of analysis, right-hand panels, show that proliferation is halted on the exposed side of the cultures with 5 uM doxorubicin but only at 100 uM drug were cells on the far side of the cultures affected. Initial experiments found that a cell cycle block occurred within 24h of drug exposure. Cultures left for longer periods exhibited continued growth from the far side of the cultures, which complicated the interpretation of data. Based on these findings the 24-hour time point was chosen as the period of incubation following exposure. Growth conditions were modified from the initial conditions used for experiments shown in FIGURE 4-3 to produce cultures which exhibited a higher level of proliferation towards the centre of the cultures and more symmetrical growth, allowing for more reliable analysis of proliferation. This was achieved by modifying the growth apparatus to allow higher and more uniform stirring and by switching oxygen levels of the atmosphere within the vessels from 5% to 20%. A target thickness of 150 um, similar to the typical maximum thickness of tumour cords observed in vivo, was chosen for MCC thickness at time of Chapter 4 - Anthracyclines 52 drug exposure. The concentration range of the four anthracyclines was narrowed to 0.3-10 u M , which was found to span the range of drug effect as was detectable via the BrdUrd-based end point. S-phase cell labelling conditions that produced a strong signal were chosen based on findings of the previous chapter (4 hour exposure at 100 u M BrdUrd). A requirement of using drug effect within closed-off M C C s as an assay for evaluation of drug penetration is that cellular proliferation must be sustained on the far side of the cultures during the period of drug exposure. FIGURE 4-4 shows that proliferation is sustained after up to a 4 hour period of closure of the cultures under the conditions chosen for the drug penetration studies. Duration of drug exposure was conservatively set to one hour to allow additional time for equilibration and wash-out procedures. N o reduction in proliferation was observed 24 hours following a 2-h period of closure (data not shown). tJ •g w ffl c .2 o IS "D fS « k_ (0 c 'E m 3 O FIGURE 4-4 Comparison of BrdUrd labelling on far-side relative near side of M C C s following 45 min and 4 h closures. Control cultures, not closed-off, were exposed to 200 u M BrdUrd for 1 h from both sides. Closed-off cultures were equilibrated for 45 min or 4 hours and then exposed to 200 u M BrdUrd from one-side for 1 hour. Data shows mean ± standard deviation (n=6). Typical results based on the finalised parameters, where M C C s were exposed for 1 hour and then incubated for 1 day to allow time for manifestation of drug effect are shown for EPI in FIGURE 4-5. Data show a gradual increase in depth of the effect of EPI when concentration is increased from 0.3 to 10 u M . Only at the 10-uM exposure is EPI found to exert an equal effect to either edge of the cultures. Cells in the central region of the M C C s show less inhibition of labelling. The M C C exposed to 10 uM EPI is approximately the same thickness it was at the time of treatment while the other cultures, exposed to lower concentrations, are seen to have grown increasingly thicker during the day following treatment. Chapter 4 - Anthracyclines 53 1-h epirubicin "O 0 OB CD 1 -t—' c 3 a . CO d 3. co a . o 1 1 5 0 I /7m v direction of drug diffusion FIGURE 4-5. Distribution of S-phase cells in MCCs 1 day after a 1-h exposure to increasing concentrations of EPI (0.3 to 10 uM) from one side. One day following the exposure, cultures were exposed to 100 u M BrdUrd for 4 hours from both sides to label S-phase cells. Cryosections counter stained with haematoxylin. Scale bars, 150 um, show approximate thickness of culture at time of drug exposure. FIGURE 4-6 shows a summary graph comparing the effect of the four anthracyclines examined on the near and the far side of M C C s 1 day following a 1-h drug exposure. BrdUrd labelling in the first 30 um of tissue on either side of the M C C s is expressed as the fraction of labelling seen on the respective sides of untreated M C C s from each experiment. For comparison, shaded regions on each panel indicate typical drug exposures achieved in humans for each drug (Rentsch et al. 1998; Pea et al. 2000; Danesi et al. 2002; Swenson et al. 2003). In all cases the relevant human exposures are not high enough to achieve an equal effect on both sides of the cultures. For human exposures, typically 70-80% of the total drug will be blood bound (Chassany et al. 1994; Chassany et al. 1996) while for the M C C work, carried out with 10% FBS, serum binding is estimated at 20-30% of total drug (Chassany et al. 1994; Chassany et al. 1996). Hence, if data were plotted versus free drug, the estimated human exposures would likely be ~3-fold lower relative to the M C C data shown in FIGURE 4-6 (assuming serum bound drug does not contribute to significantly to the tissue penetration of the drugs). For D O X , EPI and D A U , data from both near and far sides Chapter 4 - Anthracyclines 54 exhibit a general trend of decreasing labelling with increasing drug concentration. However, the MIT data show an increase in labelling on the near side with increasing drug concentration. Since MIT already exhibits a large effect at the lowest concentration, this is likely due either to non-proliferation related BrdUrd labelling or to cell loss on the near side at the higher concentrations. From the data, estimates of the decrease in drug exposure seen by cells on the far side relative to the near side were ~12-fold for DOX, ~10-fold for EPI, ~30-fold for D A U and greater than 30-fold for MIT. These values were arrived at through determination of the drug concentration which produced an effect on the near side that matched that seen at 10 uM drug on the far side. The average fraction of BrdUrd positive tissue in untreated MCCs was 0.40 ± 0.06 (SD) on near side, versus 0.46 ± 0.04 (SD) on far sides. This difference is thought to be mainly due to differences in edge detection at the tissue to growth medium transition on the near side in comparison with the more well defined tissue to membrane transition on the far side. M C C data shown in FIGURE 4-6 are expressed relative to untreated MCCs from respective experiments rather than the overall averages. Middle regions of the MCCs exhibited a significantly lower average BrdUrd stained fraction, 0.28 ± 0.05 (SD). 1-h doxorubicin 1-h epirubicin 1-h daunorubicin 1-h mitoxantrone p 0 I i I I I I * i i i i i 0.3 1 10 0.3 1 10 0.3 1 10 0.3 1 10 Drug concentration (JJM) FIGURE 4-6. Comparison of the effect of the four anthracyclines on the first 30 um of tissue on the near side (•) and far side (O) of MCCs 1 day after a 1-h exposure. Control experiments under normal (T) versus deprived conditions, low oxygen and glucose, (V) are also shown. Data show BrdUrd immunostaining expressed as the fraction of control levels seen in untreated MCCs from the same experiments. Points show mean ± SD (n = 4 - 6). Shaded boxes indicate typical clinical exposures (the concentration over a 1 h exposure that matched typical plasma AUC). Differences in the effects of the drugs on the far sides of the cultures were found to be statistically indistinguishable, P>0.05, except at the 10-uM x h exposure, P<0.005 (ANOVA), where the effect of EPI was significantly greater than DOX. Bottom panels show the relative thickness of MCCs at the end of the experiment. Data are expressed relative to MCC thickness at time of exposure, 160 ± 20 (SD) um, Data points show mean ± SD. Chapter 4 - Anthracyclines 55 Control experiments were carried out using a dual-reservoir apparatus, FIGURE 2-1C, in which both sides of M C C s were exposed to drug but with normal conditions on one side versus low oxygen and glucose on the other (see material and methods section), results also shown in FIGURE 4- 6. The low oxygen and glucose levels were chosen to simulate levels seen on the far side of the M C C s during the closed-off experiments. Results were used to estimate what effect an equal drug exposure would have had on the two sides of the M C C s during the closed-off experiments. N o significant difference in drug effect between the two sides was observed. These experiments were carried out separately from previous work and performed in different growth medium, see materials and methods, which may explain their slightly lower BrdUrd labelling values as compared with the other data presented in FIGURE 4-6. The effect of the 1-h drug treatment on M C C thickness 1 day following exposure is summarized in the lower panels of FIGURE 4-6. Data are expressed relative to the average thickness of control M C C s at the time of exposure, 160 ± 20 um (SD). Untreated cultures grown for 1 day from time of treatment reached 225 + 30 um (SD) in thickness. The thickness of drug treated cultures was found to lie within this range with M C C s exposed at the higher drug concentrations exhibiting the least amount of growth. Based on the relatively low amount of proliferation observed at the highest exposures, cell loss on the exposed side of cultures is likely no more than a few layers. In separate experiments, M C C s exposed to 100 u M D O X were only 1-2 cell layers thinner than M C C s exposed to 10 uM D O X , indicating only a modest increase in net cell loss. M C C s cultured for 3 days following clinically relevant exposures showed a similar distribution of 5- phase cells as seen after 1 day, FIGURE 4-7, (A) 3 p M x h D O X , (B) 3 p M x h EPI, (C) 1 p M x h D A U and (D) 1 p M x h MIT. Cultures continued to increase in thickness and in all cases there was an increase in proliferation towards the far sides indicating that penetration limited each drug's effectiveness. DOX EPI DAU MIT 150 pm JJ direction of drug diffusion FIGURE 4-7. Dis t r ibut ion of S-phase cells i n M C C s 3 days after a 1-h exposure from one side to the four anthracyclines at their peak human plasma levels; 3 p M D O X , 3 p M EPI, 1 p M D A U and 1 p M MIT. Three days after drug exposure, cultures were exposed to 100 p M BrdUrd for 4 hours from both sides. Cryosections were counter stained with haematoxylin. Scale bars, 150 pm, show approximate thickness of the cultures at the time of drug exposure. Chapter 4 - Anthracyclines 56 FLUORESCENCE BASED EVALUATION OF ANTHRACYCLINE DISTRIBUTION IN MCCs The profile of anthracycline-derived fluorescence within cultures was examined using the closed-off M C C apparatus immediately after the end of a 1-h drug exposure at 10 u M and 1 day follow-ing incubation in fresh medium. Only D O X , EPI and D A U , which all possess the same fluorophore, could be examined. Emission intensities of the three drugs in aqueous solution were within 10% of each other, however tissue fluorescence will be derived from a combination of emissions from free and partially quenched bound drug. FIGURE 4-8 shows average tissue fluorescence as a function of distance into the cultures from the near sides. D A U shows a much steeper gradient than D O X or EPI consistent with the effect-based data. Of the three drugs, D O X derived fluorescence accumulates the least over the 1-h exposure with EPI and D A U reaching ~1.5 and ~7 times higher levels respectively on the near sides, panels A-C. While a higher tissue accumulation of D A U was expected due to its higher rate of entry into cells a direct comparison of absolute fluorescence intensity is problematic due to changes to quantum efficiencies of drug fluorescence upon binding. Interestingly, the drugs show similar cellular accumulation from the middle regions towards the far sides of each culture, see panels D-F, which show the data on the same scale. Position within M C C s is presented as a fraction between 0 and 1 to allow comparison between M C C s taken immediately after drug exposure, which measured 140 ± 15 um in thick-ness, and M C C s incubated for the additional day, which were 185 ± 20 um in thick. Anthracycline Fluorescence following 1-h exposure Relative position within M C C FIGURE 4-8. Anthracycline fluorescence in M C C s following a 1-h exposure at 10 u M drug from one side. Red lines show data for M C C s frozen immediately after exposure and black lines show data for M C C s that were incubated for an additional day in drug free medium. Panels are scaled differently and for reference the dotted horizontal lines in each panel show the averaged peak D O X levels. Panels on the right-hand side show data on a magnified scale. Position within the cultures is scaled between 0 and 1, with 0 being the near side and 1 the far side of the M C C s . M C C s frozen immediately following exposure measured 140 ± 15 (SD) um in thickness while M C C s incubated for an additional day were 185 ± 20 (SD) um in thickness. Chapter 4 - Anthracyclines 57 A comparison of effect-based evaluation of drug distribution versus drug-derived fluorescence is shown in TABLE 4-1. Results from the two methods indicated that the effect-based data predic-tions are consistent with drug distribution as seen from fluorescence data. Effect based results were estimated from data in FIGURE 4-6. The decrease in fluorescence on the far side relative to the near side was calculated by averaging fluorescence over the first 30 pm of tissue on each side of the MCCs shown in FIGURE 4-8. Ratio of near to far side drug exposure Effect-Based assay Drug fluorescence DOX - 1 2 11 ± 5 (SD) EPI - 1 0 12 ± 5 (SD) DAU - 3 0 50 ± 5 (SD) MIT >30 TABLE 4-1. D r u g exposure to the near s ide re lat ive to the far s ide o f M C C s as es t imated f r o m the ef fect-based assay or d r u g d e r i v e d f luorescence . Effect-based results are estimated from data shown in F I G U R E 4-6, by determining the drug concentration in the medium that produced an effect on the near side that matched the effect seen on the far side at 10 pM drug, see text. The ratio of fluorescence on each side is calculated from data shown in F I G U R E 4-8 using the MCCs frozen immediately following exposure to 10 pM drug over 1 h. INHIBITION O F A C I D I C S E Q U E S T R A T I O N We decided to re-examine the effect in inhibition of acidic sequestration that was seen in the preliminary data of FIGURE 4-1 and FIGURE 4-2, in which ammonium chloride was shown to increase drug flux through cultures and reduce drug-derived tissue fluorescence but also possi-bly compromise the integrity of the MCCs themselves. The earlier work was performed using the flow-through flux apparatus, which did not allow for drug to build-up on the far side of the cultures. In addition to switching to the closed-off apparatus, the high drug concentration and long exposure period, 3-hours at 100 pM, which were thought to be responsible for histological changes in the cultures seen after daunorubicin were also reduced. Results for drug-derived fluorescence seen in the cultures immediately following 1 hour drug exposures with and without ammonium chloride are shown in FIGURE 4-9. Surprisingly, while peak tissue fluorescence is again seen to be reduced, tissue accumulation on the far sides is not seen to be increased, suggesting that inhibition of acidic sequestration of the drugs reduces accumulation but does not limit drug penetration through the tissue. It does however, improve drug distribution but achieves it mostly through the reduction in accumulation on the exposed side of the tissue rather than improved penetration to the far side. Chapter 4 - Anthracyclines 58 Effect of Ammonium Chloride in HCT-116 MCCs 1 -h exposure, 10 /vM drug 1-h daunorubicin normal + 50 mM NKCI 50 100 150 distance from edge (pm) FIGURE 4-9 Effect of inhibit ion of acid sequestration via ammonium chloride on the tissue distribution of drug derived fluorescence in HCT-116 M C C s . Cultures were frozen immediately following a 1-hour exposure to 10 u M doxorubicin, epirubicin or daunorubicin with or without 50 m M ammonium chloride. Results show fluorescence profiles observed from 10 um cryosec-tions. The effect in inhibition of acidic sequestration on proliferation following exposure to 1 u M D A U is shown in FIGURE 4-10. Results show no improvement in the ability of daunorubicin to exert an effect on the far side of the M C C following inhibition of acidic sequesterizaion. These results are consistent with published data in which inhibition of acidic sequestration by omeprazole was found to reduce cellular accumulation of doxorubicin by up to 3 fold but only produce a mar-ginal increase in flux out of cultures (Lee and Tannock 2006). 1.2 near side far side FIGURE 4-10 Effect of inhibit ion of acidic sequestration on BrdUrd labelling i n HCT-116 M C C after exposure to 1 u M daunorubicin for 1-hour from one side. 150 um thick M C C s were exposed to 1 u M daunorubicin ± 50 m M NH 4 C1 for 1 hour using the closed-off assay. Results show BrdUrd incorporation 1-day after drug exposure. Data points show mean ± SD (n=3). 4.3 DISCUSSION In this study M C C s were used in a novel configuration to isolate the role of drug penetration from other factors that determine drug efficacy with depth into tissue. By temporarily closing off Chapter 4 - Anthracyclines 59 one side of the MCCs during drug exposure a comparison could be made between a drug's effect on proliferating cells that were directly exposed to drug and those that lay on the far side of the MCCs. This technique differs from multicellular spheroid based cell survival experiments which predict overall drug toxicity with depth into tissue and combine changes in intrinsic sensitivity with penetration limited drug exposure (Durand 1990; Olive and Durand 1994). Results shown here found that limited tissue penetration of the drugs resulted in an estimated 10-30 fold difference in drug exposure between cells on the near versus far side of the cultures. None of the drugs were able to exert an equal effect on both sides of the cultures after a 1-h exposure match-ing their typical plasma AUC seen in humans, indicating penetration may limit their effectiveness. Surprisingly, in both the effect based assay and direct fluorescence assay, the drugs performed similarly on the far sides of the MCCs regardless of their differing lipophilicity and intrinsic potency and it was only on the near sides that their effects could be differentiated. This suggested that the increase in penetrative potential of the more lipophilic drugs was being countered by increased binding and sequestration within cells. Due to the sustained growth of the cultures after treatment, especially on the far side, it was not feasible to perform an in-situ cell survival assay by leaving the cultures for longer periods. However, the distribution of BrdUrd labelling after 3 days was similar to that seen 1 day after drug exposure and results were generally consistent with published work with V79 multicellular spheroids, in which cell survival assays were performed. Following DOX exposure of ~3.5 uM x h, spheroids that were incubated in drug free medium for an additional 24 hours prior to disaggregation exhibited a surviving fraction of -0.2 on the exposed outer layers versus -0.6, 130 um into the tissue (Durand 1990). In the spheroid study, reduced intrinsic sensitivity of cells in the central area of the spheroids and time of disaggregation were shown to play key roles in determining drug toxicity. In comparison with spheroids, MCCs pose a greater barrier to pene-tration due to their planar rather than spherical geometry. A main concern of closing off the MCCs during drug exposure was that deprivation of oxygen and glucose and build-up of lactate on the far (closed-off) side could potentially lead to a change in the intrinsic sensitivity of the cells to the drugs. In this study, we simulated the effect of oxygen and nutrient deprivation on the far side of the cultures in control experiments where equal drug exposure was made to both sides of MCCs but with one side under normal oxygen and glucose and the other under low oxygen and glucose. Comparison of drug effect on the two sides was found to be within experimental error for all four drugs. Results from the flux-based assay and also from direct fluorescence-based detection of the distribution of DOX, EPI and DAU in MCCs immediately after exposure were consistent with results from the effect-based assay. The tissue distribution, as determined from the effect-based assay, was found to rank in order of the rate of flux out of the drugs out from cultures. Results were also consistent with tissue distribution as observed via direct visualization of the drugs. In Chapter 4 - Anthracyclines 60 this case, the effect of fluorescent metabolites of parent drugs was expected to play only a minor role on overall tissue fluorescence because of the short period of exposure. Fluorescence in the near edge of the MCCs matched the expected ranking of drug uptake rates from reported cell suspension data (determined mostly by lipophilicity). As previously discussed, drug fluorescence seen on the far sides of the cultures was not significantly different for the three drugs. Following 1 day of wash out, drug fluorescence was much reduced on the exposed edge of the cultures but still present towards the central region of the cultures. In this study it was found that despite their differing lipophilicity and rate of entry into cells, none of the drugs performed better either in penetration to or effect on the far sides of the cultures. Results indicated that the increase in diffusion potential of the more lipophilic drugs, was tempered by the competing effects of more rapid DNA intercalation and other forms of sequestration once in the cells, as indicated by the higher level of drug-derived fluorescence and degree of drug effect on the near sides. Despite the similarities in drug effect and fluorescence seen on the far sides of the cultures, DOX and EPI did exhibit a more uniform distribution due to their lower of accumulation on the exposed edges. In general, a drug that does not distribute uniformly will require an increased plasma exposure in order to exert an effect far into tissue which will erode its tumour to normal tissue specificity. In the future, this assay could be used to identify analogues that balanced these competing effects and exhibit improved penetrative characteristics. Chapter 5 - Taxanes 61 5. TAXANES 5.1 SUMMARY Taxanes represent a class of drugs that are currently undergoing a new round of development but with little known of their ability to penetrate and distribute relative to blood vessels within solid tumours. Based on the shared mechanism of action but differing physicochemical proper-ties, we chose to compare the tissue penetration of the taxanes paclitaxel and docetaxel. Using the M C C effect-based assay, initial work was conducted to determine concentration range and waiting period required to produce an effect using the BrdUrd end-point. Experiments were then carried out to determine the dose response curves of paclitaxel and docetaxel on near and far M C C edges. Analysis of M C C effect-based data indicated that there was a 5-fold reduction for paclitaxel and 10-fold reduction for docetaxel in exposure to cells on the far sides of the cultures, indicating that both drugs exhibited poor tissue distribution. Effect-based results were compared with direct visualization of 3H-paclitaxel and 3H-docetaxel in M C C as well as tumour xenografts. In xenografts, taxanes were mapped relative to blood vessels to obtain drug profiles as a function of distance from vasculature. For M C C , cultures were exposed to stirred drug reservoirs and taxanes measured as a function of depth into tissue. Both taxanes exhibited limited penetration, with little drug reaching further than 100 um into tissue. Of the two, paclitaxel exhibited up to 2-fold better penetration than docetaxel. In xenografts, proliferation status following treatment was also assessed in relation to vasculature. U p to a 75% reduction in S-phase cells was observed in cells nearest vessels but only 50% reduction was observed in the tissue 150 um away, though in this case a reduced drug exposure to the cells far from vasculature would be expected to be only one of several factors influencing response to treatment. 5.2 RESULTS EFFECT-BASED E V A L U A T I O N O F T A X A N E TISSUE DISTRIBUTION IN M C C s . FIGURE 5-1 shows the effect of paclitaxel and docetaxel on proliferating cells in M C C s 2 days after a 1-hour exposure using the penetration apparatus shown in FIGURE 2-1B. M C C cryosections were immunostained for incorporated BrdUrd to show S-phase cells. Scale bars, 150 um, indicate the approximate thickness of the cultures at time of exposure. While the untreated cultures are seen to double in thickness over the two days from time of exposure, treated cultures show an increas-ing reduction in growth as a function of drug concentration. To estimate drug distribution, proliferation was compared on either edge of the cultures, where in untreated cultures similar proliferating fractions were observed. For both drugs at the intermediate concentration of 0.3 uM, the number of S-phase cells is clearly reduced on the exposed side of the culture while only at the 3-pM drug exposure is the S-phase fraction on the far side of the cultures seen to be reduced. Chapter 5 - Taxanes 62 1 -h paclitaxel 1 -h docetaxel 150 I direction of 11  ^drug diffusion FIGURE 5-1 S-phase cells i n HCT-116 M C C s 2 days after a 1-hour drug exposure from one side. Images show cryosections that were immunostained for incorporated BrdUrd. Arrows show direction of drug exposure. Scale bars, 150 um, show approximate thickness of cultures at time of drug exposure. Chapter 5 - Taxanes 63 Analysis BrdUrd -based penetration assay 1-h paclitaxel I I 1111 I II—I I 1111II 150 (im thick cultures Penetration assay: • exposed side o far side Control assay: * normal * deprived I I I 111II] 1 I I 11 1 I I IIIM! 2 [ • Ctrl t h i ckness 2 days later -I 01 £ 1 u 0.5 h n 1-5 h X T T " th i ckness at t ime o f exposu re 0 0.01 0.1 I I M I N I — I I 111 > 111—I I I I l l l l j Ctrl th ickness 2 days la te r - ] 5 2 [— th ickness at t ime o f exposu re -I 10 0.01 0.1 Drug (U-M) 10 FIGURE 5-2 Analysis of drug effect on the rapidly proliferating cells of the first 30 um of tissue on the near (•) and far (o) side of MCCs for paclitaxel and docetaxel. Data show BrdUrd immunostaining expressed as the fraction of levels seen in untreated MCCs. Results from control experiments under conditions of normal ( A ) versus energy deprived conditions (low oxy-gen/glucose and high lactate) ( A ) are also shown. Bottom panels show MCC thickness 2 days after treatment plotted relative to thickness at time of treatment. Points show mean ± SE (n=4-6). '*' indicates data points where the drug effect on the far side was significantly less than on the near side, p<0.005 (ANOVA). The effect of paclitaxel versus docetaxel on culture thickness was comparable for all concentrations, p>0.05 (ANOVA). FIGURE 5-2 shows results from analysis of the effect-based taxane penetration experiments. Top panels show the BrdUrd fraction in the first 30 um of tissue on the exposed versus far sides of the cultures for paclitaxel and docetaxel as a function of drug exposure. For both drugs a signifi-cantly greater reduction in proliferation is seen on the exposed side of the cultures. Comparison of the drug concentration required to produce a reduction in proliferation on the exposed side of the cultures as produced on the far side of cultures by the 3-uM exposure indicates an approxi-mate 5-fold decrease in drug exposure on the far side of the culture for paclitaxel and a ~10-fold decrease for docetaxel. Bottom panels show that despite paclitaxel exerting a more uniform effect within the cultures both drugs were similar in their ability to control culture growth, as reflected by their thickness. Control experiments were carried out to determine the effect that closing-off the culture during drug exposure had on response of cells on the far side. Conditions encoun-Chapter 5 - Taxanes 64 tered on the far side of the cultures during drug exposure were simulated using the apparatus shown in FIGURE 2-1C, in which both sides of the cultures were exposed to 0.1 u M drug but under normal conditions on the near side versus oxygen/glucose deprived conditions on the far side. Results, also shown in the top panels of FIGURE 5-2, indicate that the temporary deprivation of oxygen and nutrients on the far side of cultures during drug exposure d id not affect response to the drugs. 3 H-TAXANE DISTRIBUTION IN M C C s . To assess results from the M C C effect-based penetration assay, direct visualization of radiola-belled taxanes was carried out in M C C and tumour xenografts. Experiments were carried out using a film-based dry-autoradiographic technique described in section 2-5. Typical MCC-based results produced from the experiments are shown in FIGURE 5-3. To allow analysis, images of radiolabelled drug distribution from entire M C C cryosections were superimposed on images of the actual tissue sections, where M C C edges were clearly demarcated (see panels A -C ) . This permitted the determination of radioactivity as a function of depth into tissue. 1-h 3H-paclitaxel FIGURE 5-3 Image of 3 H -pacl i taxel distr ibution i n HCT-116 M C C fo l lowing a 1-hour exposure to 3 u M drug from one side. Top two panels show images of A | the tritium sensitive film and B | the unstained tissue cryosection. C [ M C C edges, as detected from the tissue section were overlaid onto the film image to allow computer analysis of grain density as a function of distance into the M C C . Scratches on the film layer were deleted from the image prior to analysis. Chapter 5 - Taxanes 65 FIGURE 5-4 shows the distribution of tritium labelled paclitaxel and docetaxel in M C C s following 1 & 2 hour exposures at 0.3 & 3 p M . In all cases, data show moderately lower accumulation and better penetration for paclitaxel relative to docetaxel. Increasing the drug concentration 10-fold from 0.3 p M to 3 p M resulted in only a ~3-fold increase in peak tissue levels indicating partial saturation of binding. A t the higher concentration both drugs show improved tissue penetration, indicating that accumulation of the drugs by tissue played a role in l imiting their penetration. The longer 2 h exposure time was seen to improve tissue distributions but peak tissue levels were only seen to increase for 0.3 p M docetaxel. Peak radioactivity levels observed at a depth of ~20 um into tissue was likely due to averaging over gaps in the tissue on the edge of the cultures and alignment of autoradiography images which blurred slightly on the edges. 3H-taxanes in MCC 1 hour exposure 2 hour exposure 0 50 100 150 0 50 100 150 depth into tissue (pm) FIGURE 5-4 3H-taxane distributions i n HCT-116 M C C s fo l lowing 0.3 & 3 p M exposures for 1 & 2 hours to paclitaxel ( • ) and docetaxel (•). M C C s , discs of tissue grown to 150 pm, were exposed to drug on one side with the other side closed-off to allow drug to build-up. Results are shown as a function of depth into tissue relative to the exposed side of the cultures. Points show mean ± SE (n=9). 3 H - T A X A N E D ISTRIBUTION IN T U M O U R X E N O G R A F T S . Typical data from the tumour-based 3H-taxane experiments are shown in FIGURE 5-5. Experi-ments were carried out using a film-based dry-autoradiographic technique described in section 2.5, which produced an image of radiolabelled drug on tritium sensitive film and a separate image of immunohistochemical detected tumour vasculature. The critical alignment of radioac-tive drug distribution and tissue vasculature was only possible because images of entire cryosections were captured, allowing tumour edges to be lined up on all sides. Chapter 5 - Taxanes 66 3H-paclitaxel (Tween-80), 2-h post injection,10 mg/kg i.v. -fee key: • blood vessels (CD31) • 'H-paclitaxel (xray-film) 150. urn • -:v ?..s< i s V I 4 '5* F I G U R E 5-5 Composite image of 3H-paclitaxeI distribution (black) and CD31 positive vascula-ture (red) in an HCT-116 tumour xenograft 2 hours post i.v. injection of 10 mg/kg 3H-paclitaxel. Images were obtained from tritium sensitive film clamped against 10 um tumour cryosection for 80 days, followed by immunohistochemical staining of CD31 positive tumour vasculature on the same section. Subsequent computer analysis of taxane distribution relative to vasculature was carried out with non-tumour areas and scratches in x-ray film cropped out. Chapter 5 - Taxanes 67 FIGURE 5-6 shows results from analysis of tritium labelled paclitaxel and docetaxel as a function of distance to nearest blood vessel. Data show that drug accumulation is limited to the first ~100 um of tissue, with little drug reaching 100-150 um away from vessels for both drugs. The vehicle Cremophor EL, which was expected to reduce plasma levels of free paclitaxel and increase its duration of exposure, reduced tissue drug levels at the 2-hour time point relative to Tween 80 by more than 2-fold, FIGURE 5-6. However, at the 8-hour time point paclitaxel in vehicle Tween 80 is seen to have undergone substantial washout while paclitaxel in the Cremophor EL vehicle shows increased tissue accumulation. For both vehicles the profile for paclitaxel at the 8-hour time point was found to be similar. Docetaxel in comparison is seen to fall off more sharply though in contrast to paclitaxel in Tween 80 docetaxel, also in Tween 80, showed increased accumulation at 8 hours. paclitaxel (Tween 80) 10 mg/kg i.v. paclitaxel (Cremophor EL) 10 mg/kg i.v. docetaxel (Tween 80) 5 mg/kg i.v. 2 h 8 h TTTTT 0.5-t Mil 50 100 150 0 50 100 150 0 distance to nearest vessel (/vm) FIGURE 5-6 3H-taxane distribution relative to nearest blood vessel in HCT-116 tumour xenografts. Data show distributions observed 2 and 8 hours following tail vein injection; 10 mg/kg paclitaxel with vehicle A | Tween 80 or B | Cremophor EL and 5 mg/kg docetaxel with vehicle C | Tween 80. Points show mean ± SE (n=9). COMPARISON OF XENOGRAFT AND M C C - B A S E D 3 H-TAXANE DATA. FIGURE 5-7 shows a comparison of tissue accumulation and depth of penetration of the taxanes in MCC and tumour xenografts. Peak tissue levels in the tumour xenografts were ~3-fold lower for paclitaxel and ~10-fold lower for docetaxel than the 2-hour, 0.3 uM MCC results, suggesting significantly lower free drug levels in the plasma (see Discussion). In MCCs, peak levels do not increase when going from a 1 to 2-hour exposure (p>0.05 ANOVA except for the 0.3-uM do-cetaxel). In MCCs, the depth of penetration was seen to increase with exposure. Due to the increased tissue distribution after 2 hours, total tissue radioactivity in MCCs increased moder-ately when going from 1 to 2 hours (significant differences, p<0.005 ANOVA, for 1 versus 2 hour exposures in all cases except for 0.3 uM paclitaxel). The depth of penetration was also seen to increase with reservoir drug concentrations. Given this, the xenograft-based results exhibited surprisingly large penetration half-depths in comparison with the MCC-based data given their relatively low peak tissue levels and hence free-drug concentrations. Chapter 5 - Taxanes 68 • paclitaxel (Tween 80) paclitaxel (Cremophor EL) I docetaxel (Tween 80) 314 0 12 1 1 0 1 8 re D .9 4 re 2 re 0 a> Q _ 1h 2h 3^ >M Ul 1h 2h 0.3/JM MCCs 2h 8h Tumours 150 o 3. • f lOO o ro % 50 re 1 o fl 1h 2h 3yU 1 1 1h 2h 0.3^M MCCs 2h 8h Tumours 25 50 75 penetration half-depth (pm) FIGURE 5-7 Summary of 3H-taxane peak tissue levels, total radioactivity and depth into tissue where radioactivity fal ls to hal f -maximum: paclitaxel ( • ) , docetaxel ( • ) and for tumour data paclitaxel in Cremophor E L ( ). Data show mean ± SE (n=9). Chapter 5 - Taxanes 69 EFFECT OF TAXANES O N S-PHASE CELL DISTRIBUTION IN TUMOUR XENOGRAFTS. S-phase cell distributions were mapped in tumour xenografts following drug treatment. In these experiments S-phase cells were mapped relative to tumour vasculature as a function taxane dose and time following treatment. Typical data are shown in F I G U R E 5-8. The protocol for staining for the multiple markers, BrdUrd , CD31 and pimonidazole is described in section 2-5 and F I G U R E 2-6. These experiments were conducted prior to the isotope studies and drugs were injected i.p. rather than i.v.. While images of entire tumour sections were captured, only small areas are shown in the figure. The areas show typical response to the drugs at 24-hours, both of which showed a reduction in proliferation near vessels the degree of which varied from vessel to vessel. S-phase cells 72 h following taxane treatment Untreated 72 h fol lowing treatment staining key: »' blood vessels (CD31) blood flow (DiOC7(3)) • S-phase cells (BrdUrd) hypoxic areas (Pimonidazole) paclitaxel 30 mg/kg i.p. docetaxel 20 mg /kg i.p. i •' * . '"' •'."•' ' • «' ' * - , • ; . ->• - ; ; •' ' ' ,- V,. r , • Jlj V"' -.* • -" * • - A , ••< . f ' • « » : . . * ". , * -150 um F I G U R E 5-8 S-phase cells seen i n HCT-116 tumour xenografts 72 hours fo l lowing i.p. admini-stration of paclitaxel and docetaxel at approximately their maximum tolerated doses. A general reduction in S-phase cells near vasculature was seen relative to controls, though the degree of response varied greatly between vessels. Results from analysis of entire tumour cryosections are summarized in F I G U R E 5-9. The time course of paclitaxel effect on S-phase cells 1, 3 and 6 days following treatment is shown in panel A, with raw data in the top panel and data normalized to the untreated controls in the bottom panel. Data show there is a decrease in S-phase cells over the first three days following treatment. From the normalized data, paclitaxel is seen to produce up to a 75% reduction in proliferation near vessels but only a 50% reduction in cells 150 um away. Further work looking at drug effect at the 3-day time point as a function of dose is shown in panel B for paclitaxel and panel C for docetaxel. Drug doses were chosen to span the range typically administered to mice and produce equitoxic effects for the two drugs (Nicoletti et al. 1994). From normalized data, bottom panels of B and C , a preferential effect is seen on the cells in the first -50 um of tissue away from vessels, where proliferation and drug exposures are highest. Chapter 5 - Taxanes 70 0 -I . , 1 -I , , 1 1 , , 1 0 50 100 150 0 50 100 150 0 50 100 150 distance to nearest visible vessel (um) FIGURE 5-9 Analysis of S-phase cell distribution HCT-116 tumour xenografts following i.p. administration of paclitaxel and docetaxel. A | Time course for action of 30 mg/kg paclitaxel 1, 3 and 6 days following treatment and B-C | dose response 3 days following treatment for paclitaxel, 10,15 & 30 mg/kg, and docetaxel, 5,10 & 20 mg/kg. Results are shown for raw data (top panels) and data normalized to controls (bottom panels), where a value of 1 indicates no change relative the untreated controls. Data was obtained from analysis of whole tumour cryosections containing 50-100 vessels, points show mean ± SE (n=3-5). 5.3 DISCUSSION Data from both radiolabelled drug imaging and drug effect on cell proliferation suggested poor distribution of the taxanes, with most drug limited to the first 100 um of tissue. Paclitaxel pene-trated further into tissue than docetaxel, consistent with previous reports of increased cellular accumulation of docetaxel (Lavelle et al. 1995). The greatest difference between the two drugs was seen in the MCC effect-based assay where paclitaxel showed a 2-fold increase in penetration relative to docetaxel. MCC based results for both drugs showed a partial saturation of tissue binding and improved penetration at higher concentrations. The formulation vehicle Cremophor EL had the expected effect of reducing initial peak levels of paclitaxel in tumours and increasing its duration of exposure. It was interesting that at the 8-hour time point both paclitaxel formulations produced similar drug profiles, however, the tissue AUC for 0-8 hours would still be higher for the Tween 80 data. Later time points might eventually result in equal or greater accumulation for the Cremophor EL data. Drug exposures for the MCC-based radiolabelled studies were chosen to bracket the expected mouse plasma levels. However, peak tumour levels for paclitaxel and docetaxel were respec-tively ~3 and 10-fold lower than those seen in MCC following the 0.3-uM exposure. This was Chapter 5 - Taxanes 71 despite reported plasma AUCs in mice in the range of 35 and 5 pM-hours and peak levels of 40 uM and 10 uM for paclitaxel and docetaxel respectively (Sparreboom et al. 1996; Sparreboom et al. 1998). The relatively low peak tissue accumulation seen here indicates that blood binding of the drugs plays a significant role in limiting their tissue distribution and that only the free drug is likely to contribute to tissue penetration. MCC work, which was carried out in 5% serum, likely resulted in -50% unbound drug (Song et al. 1996), compared to 2-10% unbound drug expected in vivo (Sparreboom et al. 1998; Goodman et al. 2006). Based on these findings, future work could be carried out to look at taxane penetration in MCCs at lower drug concentrations, matching the levels detected from the tumour-based work. Under these conditions, the taxanes might exhibit an even greater reduction in penetration or perhaps behave in a similar way to the 0.3-pM data. It would also be interesting to use the assay to examine the relation between exposure duration and drug concentration on tissue distribution at lower concentrations and longer exposures. The apparent depth of penetration of the drugs in tumours as compared to MCCs was greater than expected given the marked concentration dependence of tissue penetration seen there. In addition, the radial geometry observed in the HCT-116 tumour type, in which cords of tumour cells surround individual vessels (Kyle et al. 2003), should actually pose a greater barrier to penetration than does the planar disc geometry of MCCs. There are two factors which might explain the increased penetration seen in the tumour data. Firstly, there will exist unobserved vessels that lie just above or below the 10-pm tissue sections cut from the tumours. The likelihood of occurrence of unobserved vessels will increase with distance from observed ones and hence they may increase the measured penetration half-depth of a drug. Another factor that might increase the apparent penetration of the taxanes in vivo is the contribution of 3H-labeled liver metabolites, which could diffuse more easily than the parent drug. A unique property of the disc geometry of MCCs is that it presents a barrier to penetration in which there exist rapidly proliferating cells at two positions in the tissue (either edge). This symmetry made it possible to use the effect of the taxanes as an indicator of their distribution, in effect using the cells themselves as the drug measurement end-point. Results from the MCC effect-based assay were interesting in that the differential in drug effect on the two edges was not as great as might be expected from the steep drug gradients observed from the autoradiography-based data. However, the radiolabelled data produced discrete snapshots of drug distribution at two time points during drug build-up, while the effect-based assay integrated exposure over the entire process of build-up and washout. Redistribution of the drugs during washout and differ-ences in cellular retention of the drugs would affect overall response. Hence the effect-based assay probably produces a more realistic evaluation of drug penetration. The proliferation-based tumour xenograft data showed that both drugs were less effective on the cells furthest from tumour vasculature. However, it was difficult to interpret this data on its own, in terms of in vivo evidence of limited tissue penetration of the drugs because, unlike the MCC-Chapter 5 - Taxanes 72 based BrdUrd work, the cells furthest into the tissue exhibited only a fraction of the proliferation seen in those cells near vessels. Hence, the tissue furthest from vasculature will exhibit reduced sensitivity to the drugs and also experience a different lag time between drug exposure and progression through the cell cycle to G2/M block (Lopes et al. 1993). Additional factors that may influence cellular response to the drugs include indirect host toxicity, which might conceivably reduce proliferation status throughout the tumour. Based on the results of this study, the effect-based MCC assay could be used as a screening method for selecting taxane analogues for improved ability to distribute within tissue. In the development of new taxanes with improved anti-tumour capabilities, tissue penetration would be one of many parameters that could be modified. However, knowledge of tissue penetration could provide additional criteria for selecting potential analogues to further develop. Because drugs that distribute poorly will require higher driving concentrations from the blood to reach and kill the cells situated furthest from blood vessels, improving drug penetration could reduce systemic toxicity by reducing the plasma levels required to achieve maximum tumour response. In addition to increasing the concentration of drug seen by the cells furthest from blood vessels, improved drug penetration might also increase the window of exposure of these cells to thera-peutic levels of drug, which might counter the effects of the lower S-phase fraction observed in this population. Chapter 6 - Survey 73 6. S U R V E Y O F T H E P E N E T R A T I O N A N D D I S T R I B U T I O N O F S O M E C O M M O N A N T I -C A N C E R D R U G S 6.1 SUMMARY To survey the range of applicability of the MCC effect-based assay, we attempted to assess the tissue distribution of a selection of anticancer drugs from several distinct classes by comparing their effect to either side of MCCs. Drugs examined included representative agents from vinca alkaloids, antimetabolites, antibiotics, platinum agents, epipodophyllotoxins and alkylating agents. Additional data presented at the end of the chapter examines the tissue oxygenation status within MCCs using the hypoxia marker pimonidazole. 6.2 RESULTS Work presented in this chapter involves the assessment of drug distribution using the MCC effect-based assay. Cultures, 150 um thick, were grown using HCT-116 cells and penetration experiments were carried out using the apparatus described in FIGURE 2-1B in which cultures were exposed to drug for 1 h from one side. Drug exposure was followed by a 1-h wash out period after which the cultures were then returned to the growth apparatus and incubated in drug free medium for 1-3 days. Drug distribution was then inferred from the differential effect to the cells on either side of the cultures as detected using the BrdUrd based S-phase cell endpoint. V I N C A A L K A L O I D S FIGURE 6-1 shows the effect of vincristine, vinblastine and vinorelbine on proliferation in MCCs 1 day after a 1-hour exposure to one side. MCC cryosections were immunostained for incorporated BrdUrd to show S-phase cells. The cultures showed sustained proliferation in central areas even at the highest exposure, notably for vincristine and vinorelbine indicating differential response of non-cycling cells. Vincristine showed the most distinct difference between effect on the near versus far edge, while vinblastine and vinorelbine showing only marginal differences. FIGURE 6-2 shows results from analysis of the BrdUrd data from 1 day following exposure to the vinca alkaloids. Top panels show the BrdUrd fraction in the first 30 um of tissue on the near (exposed) versus far sides of the cultures as a function of drug exposure. Of the three drugs, vincristine is the only one which consistently showed a differential effect on near versus far sides, with results indicating a ~3-fold drug gradient across the culture. Chapter 6 - Survey 74 1 day after exposure vincrist ine vinblast ine v inorelbine CD -4—" c 3 a, CO o d a . co d a . CO 150 4 direct ion of drug diffusion F I G U R E 6-1 Vinca alkaloid data: S-phase cells 1 day after exposure to vincristine, vinblastine or vinorelbine for 1 hour to one side. Images show BrdUrd immunostained cryosections from HCT-116 M C C s . Scale bars, 150 pm, show approximate thickness of cultures at time of drug exposure, arrows show direction of drug diffusion. Vinblastine exhibited a greater effect on the far side than the exposed side at the two highest exposures. One possible explanation for this is that the drug was washed out from the exposed side quickly enough to rescue the cells while the cells on the far side, being closed off during the 1-hour wash out period, received a longer exposure. This explanation is consistent with the reported low retention and fast wash out of vinblastine in comparison with vincristine (Owellen et al. 1977). Another possible explanation for this reversal would be that cell loss on the exposed side at the higher exposures is giving the impression of reduced drug effect. In that case it would still be possible to estimate the drug gradient within the tissue by comparing the concentration at which their first appears to be an effect on the exposed side with the concentration required to produce a similar effect on the far side. In this case an ~3-fold gradient would be estimated. Vinorelbine, which is the most lipophilic (Rahmani-Jourdheuil et al. 1994) of the three showed no evidence for a differential effect between near and far sides. Interestingly the main difference between the drugs appears to be the effect they exert on the exposed sides of the cultures rather than the far side, where they appear quite similar. This suggests that retention/wash out of the drugs on the exposed side might be a more important difference between them than drug Chapter 6 - Survey 75 penetration. Control experiments under deprived conditions (see Materials & Methods), carried out to determine what effect closing-off the culture during drug exposure had on response of cells on the far side, showed equal response. Indicating that temporary deprivation of oxygen and nutrients on the far side of cultures during drug exposure d id not affect the findings. Analysis B r d U r d - b a s e d p e n e t r a t i o n a s s a y Penetration assay: • exposed side o far side Control assay: * normal * deprived lc^> 1 1-h vincristine 1-h vinblastine 1-h vinorelbine I 1 CD 0.5 •o 1— CO 0 1.5 i iIIIIIJ—i i 111ni| 1111 ni i—i—i—i • i • inij i MI MII| i III nil cu c 0.5 • • • • 1  i i r IIIIIJ— Ctrl thickness 1 day later T O o O O -L J -]- Ctrl thickness 1 day^later o _ T o j_ o T o J -• thickness at timcjof exposure •^^^TTnni I I IIIIIII I I llllll T T T -0 Ctrl thickness 1 day|later 1 1 1 2 i i I— thickness at time of exposure- -0.01 0.1 10 0.01 0.1 1 Drug (JJM) 10 0.01 0.1 10 FIGURE 6-2 Vinca alkaloids analysis: Analysis of drug effect on the first 30 um of tissue on the near ( • ) and far (o) side of M C C s 1 day following a 1-hour exposure to vincristine, vinblastine or vinorelbine. Data show BrdUrd immunostaining expressed as the fraction of levels seen in untreated M C C s . Results from control experiments under conditions of normal (A) versus energy deprived conditions (low oxygen and glucose) (A) are also shown. Bottom panels show M C C thickness 1 day after treatment plotted relative to thickness at time of treatment. Points show mean ± SD (n=3-4). Chap te r 6 - Su rvey 76 ANTIMETABOLITES FIGURE 6-3 s h o w s the effect of the p y r i m i d i n e ana logue gemci tab ine o n p ro l i fe ra t ion i n M C C s 1 and 3 days after a 1-hour exposure to one s ide. M C C cryosect ions we re i m m u n o s t a i n e d for incorpora ted B r d U r d to s h o w S-phase cel ls. A n u n u s u a l l y d is t inc t t rans i t ion i n pro l i fe ra t ion status due to an increase of B r d U r d labe l l i ng i n the cel ls adjacent to the ones fo rced into a b lock is seen at the 1-day time po in t . A l s o , the f irst layer of cells o n each edge behave d i f ferent ly , s h o w i n g susta ined labe l l i ng even at the h ighest exposure , than the adjacent cel ls 2-3 layers in . It m igh t be that the 1-h d r u g exposure is shor t e n o u g h that the w a s h out process is able rescue these cel ls. The 3-day data are consis tent w i t h resul ts seen 1 day after exposure w i t h the excep t ion that the cul tures at the l o w e r exposures exh ib i t increased th ickness due to sus ta ined p ro l i fe ra t ion on the far s ide. CD CO CD i— C a. CO a. o a. o CO a o gemcitabine 1 day after exposure 3 days after exposure .>-„<-•. ;,. ')':•/: &>hiW-»;i&i *.* 150 I direction of 1150 ^ drug diffusion FIGURE 6-3 G e m c i t a b i n e data : S -phase ce l l s 1 & 3 d a y s af ter e x p o s u r e to 3-90 uM g e m c i t a b i n e fo r 1 h o u r f r o m one s i d e . Images s h o w B r d U r d i m m u n o s t a i n e d c ryosec t ions f r o m H C T - 1 1 6 M C C . Scale bars , 150 u m , s h o w approx ima te th ickness of cu l tu res at time of d r u g exposure, a r rows s h o w d i rec t ion of d r u g d i f fus ion . Chapter 6 - Survey 77 FIGURE 6-4 shows results from analysis of the gemcitabine data for 1 day after exposure. Top panels show the BrdUrd fraction in the first 30 um of tissue on the near (exposed) versus far sides of the cultures as a function of drug exposure. A significantly greater reduction in proliferation is seen on the exposed side of the cultures. Comparison of the drug concentration required to produce a reduction in proliferation on the exposed side of the cultures as was produced on the far side of cultures by the 30-pM exposure indicates a >10-fold decrease in drug exposure on the far side of the culture. Bottom panel shows the effect of gemcitabine on culture thickness. Results show that the 3-uM exposure reduces culture growth to half that seen in untreated cultures, which is consistent with the BrdUrd profile shown in FIGURE 6-3. Analysis B r d l l r d - b a s e d p e n e t r a t i o n a s s a y 1-h gemci tabine Penetration assay: • exposed side o far side C *—• o > J S a CO - h-1—i—i | ± * 1 Control assay: » normal * deprived 10 30 100 1.5 cu c u \— thickness at time of exposure 0.5 h —i—i • 11 • • | 1—i—i i 1111| Ctrl thickness 1 day l a t e r M H M M H M H I 0 I I i ' ' 3 10 30 100 drug (/vM) FIGURE 6-4 Gemcitabine analysis: Analysis of drug effect on the first 30 um of tissue on the near ( • ) and far (o) side of M C C s 1 day following a 1-hour exposure to gemcitabine. Data show BrdUrd immunostaining expressed as the fraction of levels seen in untreated M C C s . Results from control experiments under conditions of normal ( A ) versus energy deprived conditions (low oxygen and glucose) (A) are also shown. Bottom panels show M C C thickness 1 day after treat-ment plotted relative to thickness at time of treatment. Points show mean ± SD (n=3-4). Chapter 6 - Survey 78 Control experiments under normal versus deprived conditions were carried out to determine the effect of closing-off the culture during drug exposure on response of cells on the far side. Results, also shown in FIGURE 6-4, showed equal response to the drug indicating that temporary depriva-tion of oxygen and nutrients on the far side of cultures during drug exposure did not affect the findings. The MCC-based work was seen to be consistent with tumour xenograft-based data from our lab also grown using HCT-116 cells (Huxham et al. 2004). Further work looking at two other pyrimidine analogues, 5-fluorouracil and cytarabine, required concentrations significantly higher than those used clinically to achieve a detectable effect. Unlike gemcitabine, both these agents are selective for S-phase cells and are commonly administered as continuous infusions over 24 hours (Goodman et al 2006). A t the concentrations used here 5-flurouracil exhibited a uniform effect on both sides and cytarabine showed a slight decrease in effect on the far side, see data FIGURE 6-5. Due to the unusually high concentrations required to produce an effect after the 1-h exposures, further work was not pursued for these two agents. 1 day after exposure CD -*—' CO CD . Z3 o o a. o o CM O O 5-fluorouracil i l a. o o CO o o o cytarabine •J''-fill;-;---;i'.vTv-- I & r * i * s * i & & & & & & 1 150 / jm direct ion of drug diffusion FIGURE 6-5 5-FU & cytarabine data: S-phase cells 1 day after exposure to 5-FU or cytarabine for 1 hour to one side. Images show BrdUrd immunostained cryosections from HCT-116 M C C . Scale bars, 150 um, show approximate thickness of cultures at time of drug exposure, arrows show direction of drug diffusion. Chap te r 6 - S u r v e y 79 P L A T I N U M AGENTS - CISPLATIN FIGURE 6-6 s h o w s the effect of c i sp la t in on pro l i fe ra t ion i n M C C s 1 & 3 d a y s after a 1-hour exposure to one s ide . M C C cryosect ions were i m m u n o s t a i n e d for i nco rpo ra ted B r d U r d to s h o w S-phase cel ls. A t the 1-day t ime po in t , c isp la t in exhib i ts inc reas ing cont ro l of M C C th ickness w i t h concent ra t ion but does no t p r o d u c e a cessat ion of B r d U r d labe l l i ng . O n l y after 3 days is labe l l ing par t ia l ly r educed at the h ighe r exposures . N o di f ference in d r u g effect o n near versus far s ides was detected. c isp la t in 1 day after exposure FIGURE 6-6 C i s p l a t i n data: S -phase ce l l s 1 & 3 days after exposu re to 10-300 fiM c i s p l a t i n for 1 h o u r f r o m one s i de . Images s h o w B r d U r d i m m u n o s t a i n e d c ryosec t ions f r o m H C T - 1 1 6 M C C . Scale bars, 150 u m , s h o w app rox ima te th ickness of cu l tures at t ime of d r u g exposure , a r rows s h o w d i rec t ion of d r u g d i f f us i on . FIGURE 6-7 s h o w s resul ts f r o m ana lys is of the B r d U r d data f r o m 3 d a y s f o l l o w i n g c isp la t in exposure. T o p pane ls s h o w the B r d U r d f ract ion i n the f irst 30 u m of t issue o n the n e a r / e x p o s e d versus far s ides of the cu l tu res as a func t ion of d r u g exposure . N o s ign i f i can t decrease in effect on Chapter 6 - Survey 80 the far side was observed. However, concentrations required to achieve a detectable reduction in BrdUrd labelling were significantly higher than what is typically encountered during human exposures, suggesting that the 1-hour exposure used here was not appropriate and that a longer exposure would likely have permitted work to be carried out at lower concentrations. Control experiments under deprived conditions were not performed for cisplatin. Analysis BrdUrd-based penetration assay 1-h cisplatin TT| I I I I M 11 [ I I I Penetration assay: • exposed side o f a r side 0 LuJ i I i i — J 10 100 Drug (uM) FIGURE 6-7 Cisplatin analysis: Analysis of drug effect on the first 30 um of tissue on the near ( • ) and far (o) side of M C C s 3 days following a 1-hour exposure to cisplatin. Data show BrdUrd immunostaining expressed as the fraction of levels seen in untreated M C C s . Bottom panels show M C C thickness 3 days after treatment plotted relative to thickness at time of treatment. Points show mean ± SD (n=3-4). Chapter 6 - Survey 81 BLEOMYCIN F I G U R E 6-8 shows the effect of bleomycin on proliferation in MCCs 1 day after a 1-hour exposure to one side. MCC cryosections were immunostained for incorporated BrdUrd to show S-phase cells. At all but the lowest exposure, bleomycin can be seen to exhibit a greater effect on the near sides of the cultures. However, bleomycin requires oxygen to exert cytotoxicity hence even with a uniform distribution it would be expected to exhibit less effect on the far side of the culture (oxygen profiles in MCCs are shown in F I G U R E 6-15). b l e o m y c i n 1 day after exposure 150 f] direction of pm " V d r u g diffusion F I G U R E 6-8 Bleomycin data: S-phase cells 1 day after exposure to bleomycin for 1 hour from one side. Images show BrdUrd immunostained cryosections from HCT-116 MCC. Scale bars, 150 um, arrows show direction of drug diffusion. F I G U R E 6-9 shows results from analysis of the BrdUrd data for 1 day following exposure to bleomycin for normal cultures, 140 um, or thick cultures, 190 um. Top panels show the BrdUrd fraction in the first 30 um of tissue on the near (exposed) versus far sides of the cultures as a function of drug exposure. Initial comparison of the effect on the near versus far sides of the normal cultures suggests a modest differential in drug exposure. However, results from the Chapter 6 - Survey 82 control/energy deprived experiments indicate that oxygen depravation on the far sides of the cultures may be completely responsible for the differential effect. For the thick cultures the difference appears to be due to a combination of drug gradient and oxygen effect. i 1 0.5 I— thickness at time of exposure \— thickness at time of exposure I 10 100 1 10 100 drug (U/L) FIGURE 6-9 Bleomycin analysis: Analysis of drug effect on the first 30 um of tissue on the near ( • ) and far (o) side of thin or thick M C C s 1 day following a 1-hour exposure to bleomycin. Data show BrdUrd immunostaining expressed as the fraction of levels seen in untreated M C C s . Results from control experiments under conditions of normal (A) versus energy deprived conditions (low oxygen and glucose) (A) are also shown. Bottom panels show M C C thickness 1 day after treatment plotted relative to thickness at time of treatment. Points show mean ± SD (n=2-3). Chapter 6 - Survey 83 ETOPOSIDE FIGURE 6-10 shows the effect of etoposide on proliferation in M C C s 1 day after a 1-hour exposure to one side. M C C cryosections were immunostained for incorporated BrdUrd to show S-phase cells. A t 3 and 10 u M , etoposide can be seen to exhibit a greater effect on the near sides of the cultures and at the highest exposure, 10 u M , etoposide exhibits a greater effect on the rapidly proliferating cells on the far side than it does in the middle region of the culture which exhibits a lower proliferating fraction. 1 day after etoposide exposure T3 CD CO 0 c =3. CO 3. o a. o o 150 / jm direct ion of drug diffusion FIGURE 6-10 Etoposide data: S-phase cells 1 day after exposure to etoposide for 1 hour from one side. Images show BrdUrd immunostained cryosections from HCT-116 M C C . Scale bars, 150 um, show approximate thickness of cultures at time of drug exposure, arrows show direction of drug diffusion. FIGURE 6-11 shows results from analysis of the BrdUrd data for 1 day following exposure to etoposide. Top panels show the BrdUrd fraction in the first 30 um of tissue on the near (exposed) versus far sides of the cultures as a function of drug exposure. Initial comparison of the effect on the near versus far sides of the cultures suggest approximately a 5-fold differential in drug Chapter 6 - Survey 84 exposure. However, results from the control/energy deprived experiments indicate that drug penetration may be only partially responsible for the difference in effect between the two sides. Based on the control studies, the gradient in drug exposure to the two sides is likely limited to ~2-fold. Bottom panel shows the effect of etoposide on culture thickness. Results show that only at 100 p M drug is there a large decrease in culture thickness relative to untreated controls. This is despite significant decreases in proliferation on both edges at lower doses. These results suggest that proliferation in the central areas of the cultures may be responsible for their sustained growth. Analysis BrdUrd-based penetration assay 1-h etoposide Penetration assay: • exposed side o far side 1.5 c u 0.5 • • • • I 1—i i 111 • • | Ctrl thickness 1 day later • thickness at time of exposure iL 10 drug (/JM) 100 FIGURE 6-11 Etoposide analysis: Analysis of drug effect on the first 30 pm of tissue on the near ( • ) and far (o) side of M C C s 1 day following a 1-hour exposure to etoposide. Data show BrdUrd immunostaining expressed as the fraction of levels seen in untreated M C C s . Results from control experiments under conditions of normal (A) versus energy deprived conditions (low oxygen and glucose and high lactate) (A) are also shown. Bottom panels show M C C thickness 1 day after treatment plotted relative to thickness at time of treatment. Points show mean ± SD (n=2-4). Chapter 6 - Survey 85 MITOMYCIN C FIGURE 6-12 shows the effect of mitomycin C on proliferation in M C C s 1 and 3 days after a 1-hour exposure to one side. M C C cryosections were immunostained for incorporated BrdUrd to show S-phase cells. A t the 1-day time point, mitomycin C exhibits partial control of M C C thickness but appears to stimulate BrdUrd labelling, more so in the central region of the cultures. After 3 days BrdUrd labelling is seen to be reduced throughout the cultures. CD CD ca i— ~*—' c 3 CO d 3. CO a. o mitomycin C 1 day after exposure 3 days after exposure . * »'. * * • l" ,"• J * V * • * > i "... • 1150 JJ, direct ion of drug diffusion FIGURE 6-12 Mitomycin C data: S-phase cells 1 & 3 days after exposure to mitomycin C for 1 hour from one side. Images show BrdUrd immunostained cryosections from HCT-116 M C C . Scale bars, 150 um, show approximate thickness of cultures at time of drug exposure, arrows show direction of drug diffusion. FIGURE 6-13 shows results from analysis of the BrdUrd data for 3 days following mitomycin C exposure. Top panels show the BrdUrd fraction in the first 30 um of tissue on the near (exposed) Chapter 6 - Survey 86 versus far sides of the cultures as a function of drug exposure. From the data, only a modest difference in drug effect is observed between the two sides. Control experiments under normal versus deprived conditions (low oxygen and glucose) showed equal drug effect. Mitomycin C did not show a differential effect in the control assay under conditions of normal versus low oxygen. In some instances mitomycin C has been reported to exhibit increased toxicity under conditions of low oxygen (Gupta and Costanzi 1987; Pan et al. 1993). One possible, though perhaps unlikely, explanation for the lack of an oxygen effect observed here would be if the drug were retained in the tissue long enough such that a significant portion of its activity occurred during the 3-day incubation period subsequent to the 1-h drug exposure, during which the tissue oxygenation on the far sides of cultures would be at normal levels. In this case, the effect-based estimate of drug distribution would remain valid. Analysis B r d U r d - b a s e d penetrat ion a s s a y 1-h mitomycin c Penetration assay: • exposed side o far side Control assay: I -» normal A deprived to C I I I I 111 1 1 I I I 1111 Ctrl thickness 3 days later \ - thickness at time of exposure -I 0.1 1 drug (p/M) 10 FIGURE 6-13 Mitomycin C analysis: Analysis of drug effect on the first 30 pm of tissue on the near ( • ) and far (o) side of M C C s 3 days following a 1-hour exposure to mitomycin C. Data show BrdUrd immunostaining expressed as the fraction of levels seen in untreated M C C s . Results from control experiments under conditions of normal (A) versus energy deprived conditions (low oxygen and glucose) (A) are also shown. Bottom panels show M C C thickness 3 days after treatment plotted relative to thickness at time of treatment. Points show mean ± SD (n=3-4). Chapter 6 - Survey 87 ALKYLATING AGENTS FIGURE 6-14 shows the effect of B C N U and thiotepa on proliferation in M C C s 3 days after a 1-hour exposure to one side. M C C cryosections were immunostained for incorporated BrdUrd to show S-phase cells. Both agents exhibited little or no effect on BrdUrd labelling or culture thickness despite exposures well above typical blood plasma levels achieved clinically. Based on the lack of detectable effect or difference in effect on either edge of the cultures further work was not pursued. 3 days after exposure 5 a: o CO bcnu . . . . . . „ .... ' ••Jf,' thiotepa 1150 £ direction of drug diffusion FIGURE 6-14 Alkylating agents data: S-phase cells 3 days after exposure to bcnu or thiotepa for 1 hour to one side. Images show BrdUrd immunostained cryosections from HCT-116 M C C . Scale bars, 150 um, show approximate thickness of cultures at time of drug exposure, arrows show direction of drug diffusion. Chapter 6 - Survey 88 H Y P O X I C P R O F I L E The hypoxia marker pimonidazole was used to define regions of low oxygen within the cultures under different gassing conditions. Pimonidazole is a 2-nitroimidazole that is reduced under conditions of low oxygen, half-maximum binding occurring at ~0.5% 0 2, and forms adducts with thiol groups in proteins, peptides and amino acids within cells (Raleigh et al. 1987; Lord et al. 1993; Azuma et al. 1997). Subsequent immunodetection of pimonidazole-thiol adducts allows visualization of regions of low oxygen within tissue. FIGURE 6-15 shows pimonidazole immu-nostaining following a 1-hour labelling period under gassing conditions of 20, 5 and 0% Oz. Results indicate that under 20% gassing, 150 pm thick cultures exhibit a small region of low oxygen along the far edge of the cultures. The hypoxic area is roughly half the culture width under 5% 0 2 gassing and the entire culture is seen to stain for pimonidazole under 0% 0 2. H y p o x i c P ro f i l e 20% 0 , 4 I 5% 0 P 150 [ I direction of pm v 0 2 diffusion FIGURE 6-15 Hypoxic profile: Pimonidazole immunostaining profile in 150 pm thick HCT-116 MCC. Cultures were equilibrated for lh hour under 20, 5 or 0% 0 2 and then exposed to 200 p M pimonidazole for 1 hour from one side. Scale bars, 150 pm, show approximate thickness of cultures at time of drug exposure, arrows show direction of pimonidazole diffusion. Chapter 6 - Survey 89 Results for analysis of a series of experiments looking at the distribution of pimonidazole staining in 150 um thick cultures under various conditions of gassing and pimonidazole concentration are shown in FIGURE 6-16. In addition to characterizing the oxygenation status of cultures during a typical drug penetration experiment, we also wanted to examine the effect of consumption of pimonidazole through tissue binding on its penetration. It was hypothesized that under condi-tions of 0% 0 2 gassing, fully hypoxic cultures might consume enough pimonidazole on the exposed side of the cultures to limit its ability to label tissue on the far side of the cultures. However, this phenomenon was not observed under the range of conditions shown in FIGURE 6-16. Staining intensity on either edge was found to be quite variable, likely due to poor control of O z status within the apparatus when under 0% 0 2 gassing. Analysis Pimonidazole binding profile FIGURE 6-16 Hypoxic profile analysis: Pimonidazole immunostaining profile following exposure to 50, 200 or 400 u M pimonidazole for 1 hour under 20, 5 or 0% O z . Points show mean ± SE (n=4). Chapter 6 - Survey 90 6.3 DISCUSSION The survey conducted here gives insight into the range of applicability and limitations in apply-ing the MCC-effect based assay to evaluate the ability of anti-cancer drugs to distribute within tissue. Overall, the 1-h exposure period followed by a 1 day incubation period and BrdUrd-based endpoint were found to work best for the vinca alkaloids, gemcitabine and etoposide. Other drugs, such as mitomycin C and cisplatin, required longer incubation periods or higher drug exposures to reduce BrdUrd labelling. The limitation in the maximum duration of sustained exposure may limit the applicability of .the assay for drugs which require extended exposures to manifest activity (e.g. cytarabine and 5-FU). Other limitations encountered here include the reduction in oxygenation on the far sides of cultures that occurs during drug exposure, which was shown through control experiments to modify the toxicity of bleomycin. While overall the BrdUrd based end point was found to yield useful results in terms of detecting drug effect, it may not be the optimal end-point for many drugs. For example cisplatin and mitomycin C required extending the delay period following drug exposure to detect an effect. Other drugs showed little effect following exposure (e.g. B C N A and thiotepa). A n apoptosis based endpoint, perhaps using the T U N E L assay, might prove more effective in some cases or possibly direct detection of drug induced D N A damage using a maker such as yH2AX. Chapter 7 -Summary & Future Directions 91 7. S U M M A R Y A N D F U T U R E D I R E C T I O N S The overall aim of this research project was to develop and validate a novel application of MCC in the assessment of the extravascular penetration and distribution of anti-cancer drugs. The original hypothesis of the study was that the symmetrical proliferation that occurs in MCC could be used as a simple means of measuring drug distribution using a biological endpoint. While there already exist several methods for evaluating drug penetration such as direct visualization of radiolabelled or fluorescent molecules and indirect flux-based experiments, none of these methods are particularly amenable as a generalised screen that could be used at an early phase in the development of new anticancer drugs. Results from this study suggest that an MCC effect-based assay could be fast and simple enough to be used to test newly synthesised drugs with little knowledge required of physicochemical parameters or drug-tissue interactions. On the most basic level the assay might be useful in determining when a drug's cellular interactions, e.g. binding, metabolism or accumulation, are significant enough to affect its ability to distribute within tissue. For example to determine what is "too much" in terms of a drug's binding site affinity. In this sense the assay could be useful in categorizing a family of related drug analogues with differing physicochemical parameters. While ideally the assay would be able to select the analogues most likely to succeed in vivo, at the very least it could provide another level of information in terms of interpreting differences in the observed anti-tumour efficacy of the drugs. An effective strategy for making use of information provided by the assay might be as a quick way of sorting related analogues so that representative compounds spanning the range of possible behaviour could be selected for in vivo testing. A summary of the MCC-based tissue distribution data presented in this report is shown in TABLE 7-1. Each drug's ability to distribute within tissue was estimated from the differential effect observed on the two sides following exposure to one side only. Drug exposure to the far sides of the cultures was seen to vary from as low as 3-10% of reservoir exposure for the anthracyclines to cases where no reduction in exposure could be detected on far sides, as was seen for 5-FU and cisplatin. In between were the taxanes, where far sides received 10-20% of reservoir exposures. Several drugs were observed to exhibit marginal reduction in exposure to the far side such as the vinca alkaloids, 33-100% of reservoir exposure. Cases where the assay was observed to fail included situations where 1 h drug exposures produced no detectable effect on the BrdUrd end-point, e.g. cytarabine and thiotepa both of which are normally administered as continuous infusions. Other examples of assay failure included drugs for which control experiments under conditions of reduced oxygen and glucose showed response on the far side to vary for reasons other than drug penetration, e.g. bleomycin and etoposide which exhibit oxygen dependent toxicity. Pertinent drug-tissue interactions and their cell cycle effects, TABLE 7-2, as well as human Chapter 7 -Summary & Future Directions 92 pharmacokinetic data and some physicochemical parameters, TABLE 7-3, have also been summa-rized. The MCC effect-based assay was found to compare reasonably well with the direct visualization of drugs in MCCs. Anthracycline gradients as predicted by the effect-based assay were found to agree with direct visualization of the drugs immediately following the 1-h exposure. Ranking of the ability of the anthracyclines to distribute within tissue by the effect-based assay was also consistent with ranking derived from flux-based data. For the taxanes, while ranking via the effect-based assay was consistent with direct visualization of the drugs, a comparison between the methods was difficult to make and it was hard to imagine how the poor drug distribution observed from radiolabelled assays would produce any effect at all to the far sides of the cultures. This apparent problem likely originates from the inherent difference in the two assays; while the effect-based assay integrates the original 1-h drug exposure as well as subsequent redistribution of drug over the following 48 h, the radiolabelled assay produces a discrete snapshot in time of Cb .c 5 c doxorubicin . epirubicin :*?diunprubicin mitoxaritrone. !MA <-ia'i:;v':»ii'ia>l MW Cone Range wait Gradient (fjM) (days) (fold) % exposure to far side Comments ref 543 543 530 444 0.3-10 0.3 - 10 0.3-10 0.3-10 12 10 30 >30 paclitaxel docetaxel 854 862 0.03 - 3 0.03 - 3 5 10 less retention than docetaxel a •j= O CO ro !2 o o .E « CO o I T3 gemcitabine 5-fluorouracil " cytarabine 263 3-90 1 >10 130 100-400 1 0 243 100- 1000 1 short exp. not relevant b short exp. not relevant b vincristine vinblastine ^Vinorelbine^ 825 811 980 0.03 - 3 0.03 - 3 0.03 - 3 3 <3 0 less retention than vincristine c more lipophilic d mitomycih;C iblepifiycin ..." cisplatin etoposide C*»£ytjiiptepai 334 1420 298 589 215 189 0.3-10 1-50 U/L 10-300 3-100 10-90 3-30 requires oxygen for effect e requires oxygen for effect e short exp not relevant b BrdUrd pimonidazole 1 - 100 50 - 400 need 100 uM for 1 h to label cells higher exp. brings up background TABLE 7-1 Summary of results obtained from the MCC effect-based assay for evaluation of drug distribution. Gradient refers to the fold reduction in exposure estimated on the far sides of MCCs following lh drug using the closed-off assay and %exposure is calculated from this. Concentration refers to the concentration range that was studied and wait refers to the incubation period following completion of drug exposure before evaluating drug effect. "(Kelland and Abel 1992; Lavelle et al. 1995), b(Goodman et al. 2006), '(Ferguson and Cass 1985),d(Rahmani-Jourdheuil et al. 1994), "(Yamauchi et al. 1987). Chapter 7 -Summary & Future Directions 93 the drug distribution as it existed immediately following the 1-h exposure. This same limitation also exists when comparing the effect-based anthracycline results with direct visualization data and it is perhaps surprising that they compared so well. A comparison between the tumour and MCC results could be made from the BrdUrd penetration work and the radiolabelled taxane work. While the distribution of BrdUrd labelled S-phase cells does not reflect actual free BrdUrd levels that existed within cells, due to the saturation in the level of BrdUrd incorporated into DNA, some insight in BrdUrd penetration can still be obtained from comparing the tumour and MCC data. MCC results found that a sustained exposure to 100 uM BrdUrd for 1 hour was required to label S-phase cells 150 um into the tissue while in tumours the 100-mg/kg BrdUrd dose did not adequately label cells 100-150 um away from vessels and was found to produce only a 30-uMxh AUC, which was consistent with the MCC prediction. The distribution of radiolabelled taxanes was found to rank similarly between MCC and tumours. The observed differences in depth of penetration of the taxanes, though reasonably w <u ,c I 5 c to CD doxorubicin epirubicin daunorubicin mitoxantrone D N A effects interaction topo II breaks binds yes yes binds yes yes binds yes yes binds yes yes microtubules cell cycle specif ic ref no no no no induced block G2 /M G2 /M G2 /M G2 /M ref paclitaxel' docetaxel stabil izes stabil izes M M M M • ± O <= -S nj ra o E ra o o .= ra > a ra ^emcitabjnes K5l1luorouracrt K.cytarabine-incorporates incorporates incorporates no S S G1 then G2 /M f S e S e • vincristine : vinblastine vinorelbine inhibits inhibits inhibits G2 /M e G2 /M e G2 /M e G2 /M G2 /M G2 /M .o w o c cn •6 mitomycin C bleomycin cisplatin etoposide ' BCNU • • thiotepa. cross links sc isson cross links cross links cross links yes yes yes yes yes no G2 no G2 /M no no S then G2 /M G2 S then G2 /M G2 /M late S G2 /M late S G2 /M BrdUrd i pimonidazole incorporates yes TABLE 7-2 Summary of D N A and microtubule interactions and cell cycle effects of the drugs tested using the effect-based assay. '(Goodman et al. 2006), b(Czyz et al. 2005), c(Durr et al. 1983), d(Blagosklonny and Fojo 1999), e(Petru et al. 1995), f(Cappella et al. 2001), g(Dorr 1992), h(Twentyman 1983), '(Erba et al. 1986), '(Korman et al. 1984). Chapter 7 -Summary & Future Directions 94 similar, were more likely due to experimental artefacts rather than differences in diffusion in the two models. The control assay, in which both edges received equal drug exposure but under different oxygen and glucose conditions, was found to work reasonably well in detecting non-diffusion related effects. Key examples were the bleomycin and etoposide studies, in which low oxygen was known to modify drug effect. It is not clear how useful this assay was in detecting acidity related effects that might occur due to decreasing pH with depth into tissue. In MCC, a drop in pH of -0.3 likely occurs at a depth of 150 um into tissue in relation to the edge, based on measurements in large, 650-pm-thick spheroids where a drop of -0.25 occurred at 150 um into the tissue (Acker et al. 1987). Hence, for experiments conducted here at pH 7.4, the far edge of the cultures likely resides at -7.1. This would be expected to produce an -25% reduction in anthracycline efficacy (Born and Eichholtz-Wirth 1981; Mahoney et al. 2003) and perhaps a 10%-15% reduction for vinca co <b "G o C <0 CO CD c 03 X to CO CO o c J D 03 B a> E CO 03 " D o O c 03 > 03 Peak p lasma, administration (pM) ref blood binding ref logP ref p K a ref active uptake ref doxorubicin ; epirubicin ' daunorubicin - mitoxantrohb 8 60 mg/m 2 , bolus a 7 90 mg/m 2 ,bolus a 0.5 45 mg/m 2 b 0.7 12 mg/m 2 , 0.5-h infusion c 7 6 % c 50 -60% i 7 7 % i 9 7 % c 0.3 j 0.6 j 0.9 j 0.8 k 8.3 n 8.1 n 8.3 n 8.1 0 ; , paclitaxel, «'•'• docetaxel 4.25 175 mg/m 2 3h inf d 2.5 85 mg/m 2 , 1.6-h infusion c 9 0 % c 9 4 % c 5.0 I 4.2 I gemcitabine =;(5-flBupr6uracil cytarabine 27 2 h inf c 10 continuous infusion c 0.4 continuous infusion c negligible c 10% c 1 3 % c -1.3 I -0.8 I -2.7 I 3.6 I 7.6 p yes c yes c yes c vincristine vinblastine vinorelbine 0.3 2 mg, bolus c 0.4 0.15 mg/kg, bolus e 1 30 mg/m 2 , 15-min infusion c low,75% c,i 50 -99% i 8 7 % c 2.6 m 3.6 m 5.8 I 7.4 m 7.5 m mitomycin C bleomycin /'cisplatin1 etoposide BCNU thiotepa 0.6 10 mg/m 2 c 0.85 U/L 15 U/m 2 , bolus d 10 100 mg/m 2 2h inf c 7 oral 50 mg/m 2 f 5 100 mg/m 2 inf 2h g 4 80 mg/m 2 h low i 9 0 % c 9 6 % c 7 7 % g low h -2.3 I -2.2 I 1.2 I 1.3 I 7.4 o BrdUrd pimonidazole 100-250 mg p 500 mg /m 2 q 8.5 8.7 r CO o I TABLE 7-3 Summary of human plasma pharmacokinetic and physicochemical properties of the drugs tested using the effect-based assay. "(Danesi et al. 1999), b(Speth et al. 1987), '(Goodman et al. 2006), d(Huizing et al. 1993), e(Zhou and Rahmani 1992), '(Toffoli et al. 2004), g(Henner et al. 1986), h(Wadler et al. 1989), '(monographs), '(Demant and Friche 1998a), k(Burns et al. 1988), '(Bank), m(Owellen et al. 1977), "(Mulder et al. 1995), "(Raghunand and Gillies 2000), p(Mahoney et al. 2003),q(Tsang et al. 2000) r(Durand and Aquino-Parsons 2006), s(Smithen 1980) Chapter 7 -Summary & Future Directions 95 alkaloids (Ferguson et al. 1984). In both cases these changes are likely too small to be detected using the assay. An underlying assumption that motivated this project was the expectation that drugs which exhibited higher levels of cellular "consumption", either through binding, sequestration or metabolism, were likely to experience more limited tissue penetration. It was therefore a surprise to find that despite the less hydrophilic anthracycline daunorubicin exhibiting significantly greater effect and higher tissue fluorescence on the exposed side of MCCs, it was still able to match the other more hydrophilic anthracyclines in effect and fluorescence observed on the far sides of the cultures. However, having said this, it still exhibited the poorest relative distribution due to its higher accumulation on the exposed edge. A similar situation may have occurred with the vinca alkaloids. The vinca alkaloids all appeared to exert a similar effect on the far sides of the cultures suggesting similar penetration but exhibited different effects on the exposed edges, likely related to differences in their accumulation and retention within cells. The taxanes on the other hand did follow the original assumption with docetaxel exhibiting poorer tissue penetra-tion to the far side of cultures in comparison with paclitaxel. While the tissue penetration of most anticancer drugs in clinical use has not been fully character-ized, it is often argued that repeated dosing regimes used in the clinic, in addition to addressing cell cycle issues also address those related to inefficient drug distribution. Under such regimes, a drug that exhibits inefficient distribution might kill tumour cells within a certain distance of the vasculature with each treatment, and work in a manner similar to peeling away layers of an onion. However, such regimes might actually provide an ideal environment for the cultivation of drug resistant phenotypes, as the cells residing away from tumour vasculature would receive repeated sub-lethal exposures the drug. Other drugs in clinical use may reach high enough levels in the tissue furthest away to exert activity despite exhibiting significant gradients away from vessels. In these cases, poor tissue penetration is actually forcing higher systemic drug exposure and eroding the original tumour to normal tissue specificity of the drug. Developing drug analogues which exhibit more uniform tissue distribution might lower the plasma levels required to achieve an effect in the tissue furthest from vessels and help preserve tumour specificity. References 96 R E F E R E N C E S : Acker, H., J. Carlsson, et al. (1987). "Influence of glucose and buffer capacity in the culture medium on growth and pH in spheroids of human thyroid carcinoma and human glioma origin." Cancer Res 47(13): 3504-8. Acker, H., G. Holtermann, et al. (1992). "Influence of glucose on metabolism and growth of rat glioma cells (C6) in multicellular spheroid culture." Int T Cancer 52(2): 279-85. Azuma, C, J. A. Raleigh, et al. (1997). "Longevity of pimonidazole adducts in spontaneous canine tumors as an estimate of hypoxic cell lifetime." Radiat Res 148(1): 35-42. BC Cancer Agency monographs, "http:/ /www.bccancer.bc.ca/HPI/DrugDatabase/ Druglndex-Pro/default.htm." Begg, A. C, N. J. McNally, et al. (1985). "A method to measure the duration of DNA synthesis and the potential doubling time from a single sample." Cytometry 6(6): 620-6. Bissery, M. C. (1995). "Preclinical pharmacology of docetaxel." Eur I Cancer 31A Suppl 4: Sl-6. Blagosklonny, M. V. and T. Fojo (1999). "Molecular effects of paclitaxel: myths and reality (a critical review)." Int T Cancer 83(2): 151-6. Born, R. and H. Eichholtz-Wirth (1981). "Effect of different physiological conditions on the action of adriamycin on Chinese hamster cells in vitro." Br T Cancer 44(2): 241-6. Brown, J. M. (1979). "Evidence for acutely hypoxic cells in mouse tumours and a possible mecha-nism of reoxygenation." Br. I. Radiol. 52: 650-656. Brown, J. M. (1993). "SR 4233 (tirapazamine): a new anticancer drug exploiting hypoxia in solid tumours." Br I Cancer 67(6): 1163-70. Burns, C. P., B. N. Haugstad, et al. (1988). "Membrane lipid alteration: effect on cellular uptake of mitoxantrone." Lipids 23(5): 393-7. Cantoni, O., P. Sestili, et al. (1990). "Comparative effects of doxorubicin and 4'-epi-doxorubicin on nucleic acid metabolism and cytotoxicity in a human tumor cell line." Cancer Chemother Pharmacol 27(1): 47-51. Cappella, P., D. Tomasoni, et al. (2001). "Cell cycle effects of gemcitabine." Int I Cancer 93(3): 401-8. Casciari, J. J., S. V. Sotirchos, et al. (1988). "Glucose diffusivity in multicellular tumor spheroids." Cancer Res 48(14): 3905-9. Chaplin, D. J., R. E. Durand, et al. (1985). "Cell selection from a murine tumour using the fluores-cent probe Hoechst 33342." Br I Cancer 51(4): 569-72. Chaplin, D. J., R. E. Durand, et al. (1986). "Acute hypoxia in tumour: Implications for modifiers of radiation effects." In. I. Radiat. Oncol. Biol. Phys. 12:1279-1282. Chaplin, D. J., P. L. Olive, et al. (1987). "Intermittent blood flow in a murine tumor: radiobiological effects." Cancer Res 47(2): 597-601. Chapman, J. D., A. J. Franko, et al. (1981). "A marker for hypoxic cells in tumours with potential clinical applicability." Br T Cancer 43(4): 546-50. Chassany, O., S. Urien, et al. (1994). "Binding of anthracycline derivatives to human serum lipoproteins." Anticancer Res 14(6A): 2353-5. Chassany, O., S. Urien, et al. (1996). "Comparative serum protein binding of anthracycline derivatives." Cancer Chemother Pharmacol 38(6): 571-3. Coley, H. M., W. B. Amos, et al. (1993). "Examination by laser scanning confocal fluorescence imaging microscopy of the subcellular localisation of anthracyclines in parent and multidrug resistant cell lines." Br I Cancer 67(6): 1316-23. Cowan, D. S. M., K. O. Hicks, et al. (1996). "Multicellular membranes as an in vitro model for extravascular diffusion in tumours." Br. I. Cancer 74: S28-S31. Crank, J. (1975). The mathematics of diffusion. Oxford, Clarendon Press. Czyz, M., A. Szulawska, et al. (2005). "Effects of anthracycline derivatives on human leukemia K562 cell growth and differentiation." Biochem Pharmacol 70(10): 1431-42. Danesi, R., P. F. Conte, et al. (1999). "Pharmacokinetic optimisation of treatment schedules for anthracyclines and paclitaxel in patients with cancer." Clin Pharmacokinet 37(3): 195-211. References 97 Danesi, R., F. Innocenti, et al. (2002). "Pharmacokinetics and pharmacodynamics of combination chemotherapy with paclitaxel and epirubicin in breast cancer patients." Br T Clin Pharma-col 53(5): 508-18. de Wolf, F. A., R. W. Staffhorst, et al. (1993). "Role of anionic phospholipids in the interaction of doxorubicin and plasma membrane vesicles: drug binding and structural consequences in bacterial systems." Biochemistry 32(26): 6688-95. Demant, E. J. and E. Friche (1998a). "Equilibrium binding of anthracycline cytostatics to serum albumin and small unilamellar phospholipid vesicles as measured by gel filtration." Bio-chem Pharmacol 55(1): 27-32. Demant, E. J. and E. Friche (1998b). "Kinetics of anthracycline accumulation in multidrug-resistant tumor cells: relationship to drug lipophilicity and serum albumin binding." Bio-chemical Pharmacology 56(9): 1209-17. Dewhirst, M. W., R. D. Braun, et al. (1998). "Temporal changes in P02 of R3230AC tumors in Fischer-344 rats." Int I Radiat Oncol Biol Phys 42(4): 723-6. Di Marco, A., A. M. Casazza, et al. (1977). "Changes of activity of daunorubicin, adriamycin and stereoisomers following the introduction or removal of hydroxyl groups in the amino sugar moiety." Chem Biol Interact 19(3): 291-302. Djordjevic, D. and W. Szybalski (1960). "Genetics of human cell lines III. Incorporation of 5-bromo and 5-iododeoxyuridine into the deoxyribonucleic acid of human cells and its effect of radiation sensitivity." journal of experimental medicine 112: 509-531. Dolbeare, F., H. Gratzner, et al. (1983). "Flow cytometric measurement of total DNA content and incorporated bromodeoxyuridine." Proceedings of the National Academy of Sciences of the United States of America 80(18): 5573-7. Dorr, R. T. (1992). "Bleomycin pharmacology: mechanism of action and resistance, and clinical pharmacokinetics." Semin Oncol 19(2 Suppl 5): 3-8. Durand, R. E. (1980). "Variable radiobiological responses of spheroids." Radiat Res 81(1): 85-99. Durand, R. E. (1982). "Use of Hoechst 33342 for cell selection from multicell systems." I Histochem Cytochem 30(2): 117-22. Durand, R. E. (1990). "Slow penetration of anthracyclines into spheroids and tumors: a therapeu-tic advantage?" Cancer Chemother Pharmacol 26(3): 198-204. Durand, R. E. and C. Aquino-Parsons (2006). "The fate of hypoxic (pimonidazole-labelled) cells in human cervix tumours undergoing chemo-radiotherapy." Radiother Oncol 80(2): 138-42. Durand, R. E. and N. E. LePard (1995). "Contribution of transient blood flow to tumour hypoxia in mice." Acta Oncologica 34(3): 317-23. Durand, R. E. and R. M. Sutherland (1972). "Effects of intercellular contact on repair of radiation damage." Exp Cell Res 71(1): 75-80. Durr, F. E., R. E. Wallace, et al. (1983). "Molecular and biochemical pharmacology of mitoxan-trone." Cancer Treat Rev 10 Supp l B: 3-11. Dvorak, H. F. (2002). "Vascular permeability factor/vascular endothelial growth factor: a critical cytokine in tumor angiogenesis and a potential target for diagnosis and therapy." T Clin Oncol 20(21): 4368-80. Dvorak, H. F., J. A. Nagy, et al. (1988). "Identification and characterization of the blood vessels of solid tumors that are leaky to circulating macromolecules." Am I Pathol 133(1): 95-109. Ensminger, W. D., S. C. Walker, et al. (1994). "Clinical pharmacology of hepatic arterial infusions of 5-bromo-2'- deoxyuridine." Cancer Res 54(8): 2121-4. Erba, E., S. Pepe, et al. (1986). "Mitozolomide activity on human cancer cells in vitro." Br I Cancer 54(6): 925-32. Ferguson, P. J. and C. E. Cass (1985). "Differential cellular retention of vincristine and vinblastine by cultured human promyelocytic leukemia HL-60/ Cl cells: the basis of differential toxic-ity." CancerRes 45(11 Pt 1): 5480-8. Ferguson, P. J., J. R. Phillips, et al. (1984). "Differential activity of vincristine and vinblastine against cultured cells." Cancer Res 44(8): 3307-12. Freyer, J. P. and R. M. Sutherland (1983). "Determination of diffusion constants for metabolites in multicell tumor spheroids." Adv Exp Med Biol 159: 463-75. References 98 Freyer, J. P., E. Tustanoff, et al. (1984). "In situ oxygen consumption rates of cells in V-79 multicel-lular spheroids during growth." I Cell Physiol 118(1): 53-61. Frezard, F. and A. Garnier-Suillerot (1990). "Comparison of the binding of anthracycline deriva-tives to purified DNA and to cell nuclei." Biochimica et Biophysica Acta 1036(2): 121-7. Frezard, F. and A. Garnier-Suillerot (1991). "Comparison of the membrane transport of anthracy-cline derivatives in drug-resistant and drug-sensitive K562 cells." European Tournal of Biochemistry 196(2): 483-91. Garcia, P., D. Braguer, et al. (1994). "Comparative effects of taxol and Taxotere on two different human carcinoma cell lines." Cancer Chemother Pharmacol 34(4): 335-43. Gatenby, R. A., E. T. Gawlinski, et al. (2006). "Acid-mediated tumor invasion: a multidisciplinary study." Cancer Res 66(10): 5216-23. Gatenby, R. A. and R. J. Gillies (2004). "Why do cancers have high aerobic glycolysis?" Nat Rev Cancer 4(11): 891-9. Gelderblom, H., J. Verweij, et al. (2002). "Influence of Cremophor El on the bioavailability of intraperitoneal paclitaxel." Clin Cancer Res 8(4): 1237-41. Giesbrecht, J. L., W. R. Wilson, et al. (1981). "Radiobiological studies of cells in multicellular spheroids using a sequential trypsinization technique." Radiation Research 86(2): 368-86. Glimelius, B., B. Norling, et al. (1988). "Extracellular matrices in multicellular spheroids of human glioma origin: increased incorporation of proteoglycans and fibronectin as compared to monolayer cultures." Apmis 96(5): 433-44. Goldacre, R. J. and B. Sylven (1962). "On the access of blood-borne dyes to various tumour regions." Br I Cancer 16: 306-22. Goodman, L. S., A. Gilman, et al. (2006). Goodman 6k Gilman's the pharmacological basis of therapeutics. New York, McGraw-Hill. Grantab, R., S. Sivananthan, et al. (2006). "The penetration of anticancer drugs through tumor tissue as a function of cellular adhesion and packing density of tumor cells." Cancer Res 66(2): 1033-9. Gratzner, H. G. (1982). "Monoclonal antibody to 5-bromo- and 5-iododeoxyuridine: A new reagent for detection of DNA replication." Science 218(4571): 474-5. Groebe, K., S. Erz, et al. (1994). "Glucose diffusion coefficients determined from concentration profiles in EMT6 tumor spheroids incubated in radioactively labeled L-glucose." Adv Exp Med Biol 361: 619-25. Groebe, K. and W. Mueller-Klieser (1996). "On the relation between size of necrosis and diameter of tumor spheroids." Int T Radiat Oncol Biol Phys 34(2): 395-401. Gupta, V. and J. J. Costanzi (1987). "Role of hypoxia in anticancer drug-induced cytotoxicity for Ehrlich ascites cells." Cancer Res 47(9): 2407-12. Harris, A. L. (2002). "Hypoxia-a key regulatory factor in tumour growth." Nat Rev Cancer 2(1): 38-47. Hashizume, H., P. Baluk, et al. (2000). "Openings between defective endothelial cells explain tumor vessel leakiness." Am T Pathol 156(4): 1363-80. Henneberry, H. P. and G. W. Aherne (1992). "Visualisation of doxorubicin in human and animal tissues and in cell cultures by immunogold-silver staining." Br I Cancer 65(1): 82-6. Henner, W. D., W. P. Peters, et al. (1986). "Pharmacokinetics and immediate effects of high-dose carmustine in man." Cancer Treat Rep 70(7): 877-80. Hicks, K. O., Y. Fleming, ef al. (1998). "Extravascular diffusion of tirapazamine: effect of metabolic consumption assessed using the multicellular layer model." Int I Radiat Oncol Biol Phys 42(3): 641-9. Hicks, K. O., S. J. Ohms, et al. (1997). "An experimental and mathematical model for the extravas-cular transport of a DNA intercalator in tumours." British lournal of Cancer 76(7): 894-903. Hicks, K. O., F. B. Pruijn, et al. (2001). "Extravascular transport of the DNA intercalator and topoisomerase poison N-[2-(Dimethylamino)ethyl]acridine-4-carboxamide (DACA): dif-fusion and metabolism in multicellular layers of tumor cells." I Pharmacol Exp Ther 297(3): 1088-98. References 99 Hicks, K. O., F. B. Pruijn, et al. (2006). "Use of three-dimensional tissue cultures to model ex-travascular transport and predict in vivo activity of hypoxia-targeted anticancer drugs." I Natl Cancer Inst 98(16): 1118-28. Hirst, D. G. and I. Denekamp (1979). "Tumour cell proliferation in relation to the vasculature." Cell Tissue Kinet 12(1): 31-42. Huizing, M. T., A. C. Keung, et al. (1993). "Pharmacokinetics of paclitaxel and metabolites in a randomized comparative study in platinum-pretreated ovarian cancer patients." I Clin Oncol 11(11): 2127-35. Hurwitz, S. J., M. Terashima, et al. (1997). "Vesicular anthracycline accumulation in'doxorubicin-selected U-937 cells: participation of lysosomes." Blood 89(10): 3745-54. Huxham, L. A., A. H. Kyle, et al. (2006). "Tirapazamine causes vascular dysfunction in HCT-116 tumour xenografts." Radiother Oncol 78(2): 138-45. Huxham, L. A., A. H. Kyle, et al. (2004). "Microregional effects of gemcitabine in HCT-116 xenografts." Cancer Res 64(18): 6537-41. Jain, R. J. (1994). "Barriers to drug delivery in solid tumours." Sci. Amer. Tuly: 59-65. Jain, R. K. (1987). "Transport of molecules in the tumor interstitium: a review." Cancer Research 47(12): 3039-51. Jain, R. K. (1997). "Delivery of molecular and cellular medicine to solid tumors." Adv Drug Deliv Rev 26(2-3): 71-90. Janssen, H. L., K. M. Haustermans, et al. (2005). "Hypoxia in head and neck cancer: how much, how important?" Head Neck 27(7): 622-38. Kelland, L. R. and G. Abel (1992). "Comparative in vitro cytotoxicity of taxol and Taxotere against cisplatin-sensitive and -resistant human ovarian carcinoma cell lines." Cancer Chemother Pharmacol 30(6): 444-50. Kerr, D. J., T. E. Wheldon, et al. (1988). "Cytotoxic drug penetration studies in multicellular tumour spheroids." Xenobiotica 18(6): 641-8. Kimura, H., R. D. Braun, et al. (1996). "Fluctuations in red cell flux in tumor microvessels can lead to transient hypoxia and reoxygenation in tumor parenchyma." Cancer Res 56(23): 5522-8. Kinsella, T. J., J. B. Mitchell, et al. (1984a). "Continuous intravenous infusions of bromodeoxyu-ridine as a clinical radiosensitizer." T Clin Oncol 2(10): 1144-50. Kinsella, T. J., A. Russo, et al. (1984b). "A Phase I study of intermittent intravenous bromodeoxyu-ridine (BUdR) with conventional fractionated irradiation." International Tournal of Radiation Oncology, Biology, Physics 10(1): 69-76. Knemeyer, I., M. G. Wienties, et al. (1999). "Cremophor reduces paclitaxel penetration into bladder wall during intravesical treatment." Cancer Chemother Pharmacol 44(3): 241-8. Korman, D. B., T. V. Krutova, et al. (1984). "The effect of thiophosphamide on the proliferative activity of human tumor cells cultured in vivo in diffusion chambers." Arch Gesch-wulstforsch 54(2): 105-11. Krogh, A. (1919). "The number and distribution of capillaries in muscles with calculations of the oxygen pressure head necessary for supplying the tissue." I Physiol 52(6): 409-415. Kuh, H. J., S. H. Jang, et al. (1999). "Determinants of paclitaxel penetration and accumulation in human solid tumor." I Pharmacol Exp Ther 290(2): 871-80. Kyle, A. (1999). Modelling extravascular drug penetration using multilayered cell culture. Department of Physics. Vancouver, Univeristy of British Columbia. M.Sc: 133. Kyle, A. H., L. A. Huxham, et al. (2003). "Tumor distribution of bromodeoxyuridine-labeled cells is strongly dose dependent." Cancer Res 63(18): 5707-11. Kyle, A. H. and A. I. Minchinton (1999). "Measurement of delivery and metabolism of tira-pazamine to tumour tissue using the multilayered cell culture model." Cancer Chemother Pharmacol 43(3): 213-20. Lankelma, J., H. Dekker, et al. (1999). "Doxorubicin gradients in human breast cancer." Clin Cancer Res 5(7): 1703-7. Lavelle, F., M. C. Bissery, et al. (1995). "Preclinical evaluation of docetaxel (Taxotere)." Semin Oncol 22(2 Suppl 4): 3-16. References 100 Lee, C. M. and I. F. Tannock (2006). "Inhibition of endosomal sequestration of basic anticancer drugs: influence on cytotoxicity and bssue penetration." Br I Cancer 94(6): 863-9. Leighton, J. (1951). "A sponge matrix method for tissue culture; formation of organized aggre-gates of cells in vitro." T Natl Cancer Inst 12(3): 545-61. Li, C , R. A. Newman, et al. (2000). "Biodistribution of paclitaxel and poly(L-glutamic acid)-paclitaxel conjugate in mice with ovarian OCa-1 tumor." Cancer Chemother Pharmacol 46(5): 416-22. Liedert, B., D. Pluim, et al. (2006). "Adduct-specific monoclonal antibodies for the measurement of cisplatin-induced DNA lesions in individual cell nuclei." Nucleic Acids Res 34(6): e47. Ljungkvist, A. S., J. Bussink, et al. (2006). "Dynamics of hypoxia, proliferation and apoptosis after irradiation in a murine tumor model." Radiat Res 165(3): 326-36. Lopes, N. M., E. G. Adams, et al. (1993). "Cell kill kinetics and cell cycle effects of taxol on human and hamster ovarian cell lines." Cancer Chemother Pharmacol 32(3): 235-42. Lord, E. M., L. Harwell, et al. (1993). "Detection of hypoxic cells by monoclonal antibody recogniz-ing 2-nitroimidazole adducts." Cancer Res 53(23): 5721-6. Mahoney, B. P., N. Raghunand, et al. (2003). "Tumor acidity, ion trapping and chemotherapeutics. I. Acid pH affects the distribution of chemotherapeutic agents in vitro." Biochem Phar-macol 66(7): 1207-18. Martin, C, J. Walker, et al. (2003). "The expression of P-glycoprotein does influence the distribu-tion of novel fluorescent compounds in solid tumour models." Br I Cancer 89(8): 1581-9. Minchinton, A. I., R. E. Durand, et al. (1990). "Intermittent blood flow in the KHT sarcoma - flow cytometry studies using Hoechst 33342." Br. T. Cancer 62:195-200. Minchinton, A. I. and I. F. Tannock (2006). "Drug penetration in solid tumours." Nat Rev Cancer 6(8): 583-92. Minchinton, A. I., K. R. Wendt, et al. (1997). "Multilayers of cells growing on a permeable support: An in vitro tumour model." Acta Oncologica 36: 13-16. Moulder, J. E. and S. Rockwell (1987). "Tumor hypoxia: its impact on cancer therapy." Cancer Metastasis Rev 5(4): 313-41. Mueller-Klieser, W., J. P. Freyer, et al. (1986). "Influence of glucose and oxygen supply conditions on the oxygenation of multicellular spheroids." Br I Cancer 53(3): 345-53. Mulder, H. S., H. Dekker, et al. (1995). "The P-glycoprotein-mediated relative decrease in cytosolic free drug concentration is similar for several anthracyclines with varying lipophilicity." Biochem Pharmacol 50(7): 967-74. Nederman, T., J. Carlsson, et al. (1988). "Penetration of substances into tumour tissue. Model studies using saccharides, thymidine and thymidine-5'-triphosphate in cellular sphe-roids." Cancer Chemother Pharmacol 22(1): 21-5. Nederman, T., J. Carlsson, et al. (1981). "Penetration of substances into tumor tissue—a methodo-logical study on cellular spheroids." In Vitro 17(4): 290-8. Nicholson, K. M., M. C. Bibby, et al. (1997). "Influence of drug exposure parameters on the activity of paclitaxel in multicellular spheroids." Eur I Cancer 33(8): 1291-8. Nicoletti, M. I., V. Lucchini, et al. (1994). "Comparison of paclitaxel and docetaxel activity on human ovarian carcinoma xenografts." Eur I Cancer 30A(5): 691-6. Olive, P. L., D. J. Chaplin, et al. (1985). "Pharmacokinetics, binding and distribution of Hoechst 33342 in spheroids and murine tumours." Br I Cancer 52(5): 739-46. Olive, P. L. and R. E. Durand (1994). "Drug and radiation resistance in spheroids: cell contact and kinetics." Cancer Metastasis Rev 13(2): 121-38. Overgaard, J., H. S. Hansen, et al. (1989). "Misonidazole combined with split-course radiotherapy in the treatment of invasive carcinoma of larynx and pharynx: report from the DAHANCA 2 study." Int I Radiat Oncol Biol Phys 16(4): 1065-8. Overgaard, J., H. S. Hansen, et al. (1998). "A randomized double-blind phase III study of nimora-zole as a hypoxic radiosensitizer of primary radiotherapy in supraglottic larynx and pharynx carcinoma. Results of the Danish Head and Neck Cancer Study (DAHANCA) Protocol 5-85." Radiother Oncol 46(2): 135-46. References 101 Owellen, R. J., D. W. Donigian, et al. (1977). "Correlation of biologic data with physico-chemical properties among the vinca alkaloids and their congeners." Biochem Pharmacol 26(13): 1213-9. Ozols, R. F., G. Y. Locker, et al. (1979). "Pharmacokinetics of adriamycin and tissue penetration in murine ovarian cancer." Cancer Res 39(8): 3209-14. Pan, S. S., F. Yu, et al (1993). "Enzymatic and pH modulation of mitomycin C-induced DNA damage in mitomycin C-resistant HCT 116 human colon cancer cells." Mol Pharmacol 43(6): 870-7. Patterson, L. H. and S. R. McKeown (2000). "AQ4N: a new approach to hypoxia-activated cancer chemotherapy." Br I Cancer 83(12): 1589-93. Pea, F., D. Russo, et al. (2000). "Liposomal daunorubicin plasmatic and renal disposition in patients with acute leukemia." Cancer Chemother Pharmacol 46(4): 279-86. Peschke, P., V. Klein, et al. (1999). "Morphometric analysis of bromodeoxyuridine distribution and cell density in the rat Dunning prostate tumor R3327-AT1 following treatment with ra-diation and /or hyperthermia." Histol Histopathol 14(2): 461-9. Peterson, H. I. and L. Appelgren (1977). "Tumour vessel permeability and transcapillary ex-change of large molecules of different size." Bibl Anat(15 Pt 1): 262-5. Petru, E., J. Haas, et al. (1995). "[Do cell cycle changes of human ovarian carcinoma cells after exposure to cytostatic drugs in vitro correlate with cytotoxicity?]." Gynakol Geburtshilfli-che Rundsch 35 Suppl 1:15-8. Phillips, R. M., P. M. Loadman, et al. (1998). "Evaluation of a novel in vitro assay for assessing drug penetration into avascular regions of tumours." Br T Cancer 77(12): 2112-9. Phillips, T. L., V. A. Levin, et al. (1991). "Evaluation of bromodeoxyuridine in glioblastoma multiforme: a Northern California Cancer Center Phase II study." Int I Radiat Oncol Biol Phys 21(3): 709-14. Primeau, A. J., A. Rendon, et al. (2005). "The distribution of the anticancer drug Doxorubicin in relation to blood vessels in solid tumors." Clin Cancer Res 11(24 Pt 1): 8782-8. Raghunand, N. and R. J. Gillies (2000). "pH and drug resistance in tumors." Drug Resist Updat 3(1): 39-47. Rahmani-Jourdheuil, D., F. Coloma, et al (1994). "Human hepatic uptake of two vinca alkaloids: navelbine and vincristine." T Pharm Sci 83(4): 468-71. Raleigh, J. A., G. G. Miller, et al. (1987). "Fluorescence immunohistochemical detection of hypoxic cells in spheroids and tumours." Br I Cancer 56(4): 395-400. Rentsch, K. M., R. A. Schwendener, et al. (1998). "Pharmacokinetic studies of mitoxantrone and one of its metabolites in serum and urine in patients with advanced breast cancer." Eur T Clin Pharmacol 54(1): 83-9. Rew, D. A., I. D. Campbell, et al. (1992). "Proliferation indices of invasive breast carcinomas after in vivo 5-bromo-2'-deoxyuridine labelling: a flow cytometric study of 75 tumours." Brit-ish lournal of Surgery 79(4): 335-9. Rodriguez, R., M. A. Ritter, et al. (1994). "Kinetics of cell labeling and thymidine replacement after continuous infusion of halogenated pyrimidines in vivo." Int T Radiat Oncol Biol Phys 29(1): 105-13. Rubin, P. and G. Casarett (1966). "Microcirculation of tumors - Part I: Anatomy, Function and necrosis." Clin. Radiol. 17: 220-229. Sham, E. and R. E. Durand (1999). "Repopulation characteristics and cell kinetic parameters resulting from multi-fraction irradiation of xenograft tumors in SCID mice." International lournal of Radiation Oncology, Biology, Physics 43(3): 617-22. Smallbone, K., R. A. Gatenby, et al. (2006). "Metabolic changes during carcinogenesis: Potential impact on invasiveness." I Theor Biol. Smithen, C , Clarke, E., Dale, J., Jacobs, R., Wardman, P., Watts, M., and Woodcock, M. (1980). Novel (nitio-l-imidazoyl)-alkanolamines as potential radiosensitizers with improved therapeutic properties. Radiation Sensitizers. Their Use in the Clinical Management of Cancer. L. Brady. New York, Masson: 22-32. References 102 Song, D v L. F. Hsu, et al. (1996). "Binding of taxol to plastic and glass containers and protein under in vitro conditions." T Pharm Sci 85(1): 29-31. Song, D., M. G. Wientjes, et al. (1997). "Bladder tissue pharmacokinetics of intravesical taxol." Cancer Chemother Pharmacol 40(4): 285-92. Sparreboom, A., O. van Tellingen, et al. (1996). "Nonlinear pharmacokinetics of paclitaxel in mice results from the pharmaceutical vehicle Cremophor EL." Cancer Res 56(9): 2112-5. Sparreboom, A., O. van Tellingen, et al. (1998). "Preclinical pharmacokinetics of paclitaxel and docetaxel." Anticancer Drugs 9(1): 1-17. Speth, P. A., P. C. Linssen, et al. (1987). "Leukemic cell and plasma daunomycin concentrations after bolus injection and 72 h infusion." Cancer Chemother Pharmacol 20(4): 311-5. Sung, C , R. L. Dedrick, et al. (1993). "The spatial distribution of immunotoxins in solid tumors: assessment by quantitative autoradiography." Cancer Res 53(9): 2092-9. Sutherland, R. M. (1988). "Cell and environment interactions in tumor microregions: the multicell spheroid model." Science 240(4849): 177-84. Sutherland, R. M. and R. E. Durand (1976). "Radiation response of multicell spheroids~an in vitro tumour model." Curr Top Radiat Res O 11(1): 87-139. Sutherland, R. M., W. R. Inch, et al. (1970). "A multi-component radiation survival curve using an in vitro tumour model." Int I Radiat Biol Relat Stud Phys Chem Med 18(5): 491-5. Swabb, E. A., J. Wei, et al. (1974). "Diffusion and convection in normal and neoplastic tissues." Cancer Res 34(10): 2814-22. Swenson, C. E., L. E. Bolcsak, et al. (2003). "Pharmacokinetics of doxorubicin administered i.v. as Myocet (TLC D-99; liposome-encapsulated doxorubicin citrate) compared with conven-tional doxorubicin when given in combination with cyclophosphamide in patients with metastatic breast cancer." Anticancer Drugs 14(3): 239-46. Tannock, I. F. (1968). "The relation between cell proliferation and the vascular system in a transplanted mouse mammary tumour." Br I Cancer 22(2): 258-73. Tannock, I. F. (2001). "Tumor physiology and drug resistance." Cancer Metastasis Rev 20(1-2): 123-32. Tannock, I. F., C. M. Lee, et al. (2002). "Limited penetration of anticancer drugs through tumor tissue: a potential cause of resistance of solid tumors to chemotherapy." Clin Cancer Res 8(3): 878-84. Tannock, I. F. and D. Rotin (1989). "Acid pH in tumors and its potential for therapeutic exploita-tion." Cancer Res 49(16): 4373-84. ten Tije, A. J., J. Verweij, et al. (2003). "Pharmacological effects of formulation vehicles : implica-tions for cancer chemotherapy." Clin Pharmacokinet 42(7): 665-85. Terheggen, P. M., B. G. Floot, et al. (1987). "Immunocytochemical detection of interaction products of cis-diamminedichloroplatinum(II) and cis-diammine(l,l-cyclobutanedicarboxylato)platinum(ll) with DNA in rodent tissue sections." Cancer Res 47(24 Pt 1): 6719-25. Thomlinson, R. H. and L. H. Gray (1955). "The histological structure of some human lung cancers and the possible implications for radiotherapy." Br. I. Cancer 9: 539-549. Tilby, M. J., C. Johnson, et al. (1991). "Sensitive detection of DNA modifications induced by cisplatin and carboplatin in vitro and in vivo using a monoclonal antibody." Cancer Res 51(1): 123-9. Toffoli, G., G. Corona, et al. (2004). "Pharmacokinetic optimisation of treatment with oral etoposide." Clin Pharmacokinet 43(7): 441-66. Tsang, R. W., A. W. Fyles, et al. (2000). "Interrelationship of proliferation and hypoxia in carci-noma of the cervix." International Tournal of Radiation Oncology. Biology, Physics 46(1): 95-9. Tunggal, J. K., T. Melo, et al. (2000). "The influence of expression of P-glycoprotein on the penetra-tion of anticancer drugs through multicellular layers." Int I Cancer 86(1): 101-7. Twentyman, P. R. (1983). "Bleomycin—mode of action with particular reference to the cell cycle." Pharmacol Ther 23(3): 417-41. University of Alberta Drug Bank, http:/ / redpoll.pharmacy.ualberta.ca/drugbank/. References 103 Urtasun, R. C, C. J. Koch, et al. (1986). "A novel technique for measuring human tissue p02 at the cellular level." Br T Cancer 54(3): 453-7. Venkatasubramanian, R., M. A. Henson, et al (2006). "Incorporating energy metabolism into a growth model of multicellular tumor spheroids." T Theor Biol 242(2): 440-53. Vescio, R. A., C. H. Redfern, et al. (1987). "In vivo-like drug responses of human tumors growing in three-dimensional gel-supported primary culture." Proc Natl Acad Sci U S A 84(14): 5029-33. Wadler, S., M. J. Egorin, et al. (1989). "Phase I clinical and pharmacokinetic study of thiotepa administered intraperitoneally in patients with advanced malignancies." T Clin Oncol 7(1): 132-9. Walenta, S., J. Doetsch, et al. (2000). "Metabolic imaging in multicellular spheroids of oncogene-transfected fibroblasts." T Histochem Cytochem 48(4): 509-22. West, G. W., R. Weichselbaum, et al. (1980). "Limited penetration of methotrexate into human osteosarcoma spheroids as a proposed model for solid tumor resistance to adjuvant che-motherapy." Cancer Res 40(10): 3665-8. Wilson, G. D., N. J. McNally, et al. (1985). "The labelling index of human and mouse tumours assessed by bromodeoxyuridine staining in vitro and in vivo and flow cytometry." Cy-tometry 6(6): 641-7. Yamauchi, T., T. A. Raffin, et al. (1987). "Differential protective effects of varying degrees of hypoxia on the cytotoxicities of etoposide and bleomycin." Cancer Chemother Pharmacol 19(4): 282-6. Yang, J., J. Richards, et al. (1979). "Sustained growth and three-dimensional organization of primary mammary tumor epithelial cells embedded in collagen gels." Proc Natl Acad Sci U S A 76(7): 3401-5. Zackrisson, B., P. Flygare, et al. (2002). "Cell kinetic changes in human squamous cell carcinomas during radiotherapy studied using the in vivo administration of two halogenated pyrimidines." European Tournal of Cancer 38(8): 1100-6. Zhou, X. J., M. Placidi, et al (1994). "Uptake and metabolism of vinca alkaloids by freshly isolated human hepatocytes in suspension." Anticancer Res 14(3A): 1017-22. Zhou, X. J. and R. Rahmani (1992). "Preclinical and clinical pharmacology of vinca alkaloids." Drugs 44 Suppl 4:1-16; discussion 66-9. Zunino, F., A. Di Marco, et al. (1980). "The interaction of daunorubicin and doxorubicin with DNA and chromatin." Biochim Biophys Acta 607(2): 206-14. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100652/manifest

Comment

Related Items