Open Collections

UBC Undergraduate Research

Characterizing Tumour Vessels using MRI and Histology - A novel dual injection MR protocol to study tumour.. 2009

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
Moosvi_Firas_UBC_2009_Physics_449_Honours_Thesis.pdf [ 6.89MB ]
Metadata
JSON: 1.0085960.json
JSON-LD: 1.0085960+ld.json
RDF/XML (Pretty): 1.0085960.xml
RDF/JSON: 1.0085960+rdf.json
Turtle: 1.0085960+rdf-turtle.txt
N-Triples: 1.0085960+rdf-ntriples.txt
Citation
1.0085960.ris

Full Text

Characterizing Tumour Vessels using MRI and Histology A novel dual injection MR protocol to study tumour blood vessel permeability. by Firas Moosvi A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF HON. BACHLEOR OF SCIENCE in The Faculty of Graduate Studies (BioPhysics) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2009 c© Firas Moosvi 2009 Abstract Galbumin, an MR contrast agent is characterized for use in a new class of animal MR experi- ments. It’s suitability as both a T1 and T ∗2 agent was assessed and it was found that although Galbumin’s relaxivity (4.33 to 5.77 (mM · sec)−1 was comparable to Gd-DTPA, the solution was not available at a high enough concentration to achieve similar T1 weighted effects. Further, it was deemed an unworthy candidate for T ∗2 -weighted imaging as it’s magnetic susceptibility was much too low (2.95 ppm/mM). Finally, we established a theoretical basis for a novel dual contrast agent MR protocol to investigate blood vessel permeability, extracted from previously published work [1] on modelling MR contrast agents. The over-arching goal of this study is to use the live imaging capabilities of MR combined with traditional immunohistochemical techniques to more accurately characterize tumour vessel permeability. Firas Moosvi fmoosvi@interchange.ubc.ca ii Table of Contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 Introduction, Motivation and Theory . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1 Introduction and Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 Theory - Blood vessel and tumour biology . . . . . . . . . . . . . . . . . . . . . . 3 1.2.1 Blood Vessels, Extravasation and drug delivery . . . . . . . . . . . . . . . . 4 1.2.2 Immunohistochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.3 MR Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.3.1 Measuring T1 Relaxation (using a Spin Echo Pulse Sequence) . . . . . . . 9 1.3.2 Measuring T ∗2 Relaxation using (CPMG Pulse Sequence) . . . . . . . . . . 11 1.4 MR Contrast Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.5 Modelling Contrast Agents - Kety Model . . . . . . . . . . . . . . . . . . . . . . . 15 2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.1 Mouse Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.1.1 Preparing a mouse for an MR scan . . . . . . . . . . . . . . . . . . . . . . 18 2.2 Histology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.3 MR Scans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.3.1 Diffusion-weighted EPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.3.2 T1 weighted images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.3.3 T∗2 weighted images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.4 Galbumin Relaxivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 iii Table of Contents 3 Results, Analysis and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.1 Galbumin Histology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.2 Galbumin as T1 agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.3 ADC (Apparent Diffusion Constant) Maps . . . . . . . . . . . . . . . . . . . . . . 30 4 T ∗2 Weighted Imaging: Galbumin & Feridex . . . . . . . . . . . . . . . . . . . . . 33 4.1 Galbumin as a T ∗2 agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.2 Feridex as a T ∗2 agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.3 Mapping Feridex Histologically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 4.3.1 Chemical Stain - Prussian Blue . . . . . . . . . . . . . . . . . . . . . . . . 37 5 Conclusions, Implications and Future Work . . . . . . . . . . . . . . . . . . . . . 38 5.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 5.2 Implications and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Appendices A Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 A.1 Susceptibility Equation Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 A.2 IDL Program to Calculate DSC Parameter Maps . . . . . . . . . . . . . . . . . . . 44 A.3 Sample Staining Worksheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 A.4 Feridex Staining Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 iv List of Tables 1.1 A table of the typical divisions of MR contrast agents [2]. In this study, we use Feridex, Gd-DTPA and Galbumin (mainly a T1 agent). Note, the table doesn’t quantify the difference between “significant” and “dramatic”, because there is a range of effects depending on concentration and relaxivities. . . . . . . . . . . . . . 15 1.2 EES refers to the space into which tracer can leak from a capillary and Ktrans is the volume transfer constant between blood plasma and the EES. Table parameters by: [1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 4.1 Table parameters from: [2] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 v List of Figures 1.1 On the left, is a schematic of normal tissue vasculature. Notice in particular, approximate symmetry and order compared to the cancerous tissue, where the vasculature is non-uniform, inefficient and wasteful (in terms of nutritive delivery). Note also in the background, the growth of ordered layers of normal cell on the left vs. the rapidly dividing cancer cells on the right. Image credit: [3] . . . . . . . 2 1.2 Histology analysis - image of a tumour section where individual cells can be re- solved. The colours on the tumour are explained in the legend on the right but the main idea here is that around active blood vessels (stained dark blue), there is cell growth. Cells close to the blood vessels, are more likely to receive nutrients and are therefore actively dividing. The green rim, no less than 150 µm away from blood vessels, represents cells that lack oxygen because it doesn’t diffuse that far. Above is an overlay of multiple images imaged at different fluorescent wavelengths. Data credit: Alastair Kyle (AIM lab), BCCRC . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Image of a tumour frozen in an OCTmedium being sectioned at a cryostat machine. 10µm sections adhere to glass microscope slides due to a temperature gradient. . . 6 1.4 A schematic representation of the Galbumin molecule. 15-20 Gd-DTPA atoms surround a large 74 kDa Albumin protein and 3-5 fluorescent FITC tags allow for convenient histological imaging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.5 On the left, a typical proton is shown with its magnetization vector pointing along the z-axis. On the right, the axes are rotated so the schematic is showing the proton’s spin vector, projected along the x-y-axis. . . . . . . . . . . . . . . . . . . 7 1.6 Classical diagram showing spins lining up to the external magnetic field B0 re- sulting in a net magnetization M0. The spins align parallel and anti-parallel due to a small difference in energy, as shown by the energy splitting diagram (due to Zeeman splitting). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.7 Schematic representation of a Spin-Echo pulse sequence. Multiple iterations of the spin-echo sequence (180◦ then 90◦) pulses are performed and the echo is recorded as a function of time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 vi List of Figures 1.8 The signals (from figure 1.7) are Fourier transformed and here, the max signal intensity (of each echo) is plotted as a function of the variable delay. The time constant from the exponential recovery is T1 . More precisely, T1 is measured as the time it takes for the signal intensity to return to 63 % of its’ maximum value. 10 1.9 Schematic representation of a CPMG pulse sequence. The initial 90◦ pulse tips the magnetization pulse in the transverse direction, the magnetization then decays rapidly due to field inhomogeneities and other spin-spin interactions. Application of a 180◦ pulse flips the direction of the dephasing spins and causes them to rephase. As the spins rephase, a spin echo signal is generated. The T2 decay curve in blue is represented in figure 1.10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.10 The curve above shows the T2 relaxation process. The FID signal from figure 1.9 is fourier transformed and the maximum intensity is plotted as a function of the Echo Time (TE). The decay constant is the spin-spin relaxation rate. . . . . . . . 12 1.11 On the left, the magnetization vector tipped and as it returns back to B0, it precesses around B0. [4]. On the right, a schematic of slice-selection with a patient lying along the a magnetic field along their length. . . . . . . . . . . . . . . . . . . 13 1.12 Left: representation of the FID signal data acquired, notice concentration of data around the origin. Right : Typical MR image of a tumour after a Fourier Transform. Bright bulb at the bottom is the crosssection of a vial of water, used as a standard reference point. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.13 Two compartment Kety Model with rate constants describing flow of contrast agent out into the EES (compartment A)and back into the vessels (compartment B). . . 16 1.14 Schematic of a dual contrast agent protocol - the first contrast agent is injected, the signal acquired and then the second agent is administered. Because both sets of data have been acquired, the signal intensities can be processed and the two processes can be treated as independent. . . . . . . . . . . . . . . . . . . . . . . . 17 2.1 Description of Experimental setup . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.1 Typical tumour (HCT116) section stained for blood vessels (red, CD31) and Gal- bumin (green, FITC). Sections of the same tumour are stained and (fluorescently) imaged separately and overlaid on top of each other. . . . . . . . . . . . . . . . . 25 3.2 An area of the tumour above (figure 3.1) zoomed in. We note some vessels (red, CD31- bottom left) have positive Galbumin staining (green, FITC) while other vessels show no positive staining (top right). . . . . . . . . . . . . . . . . . . . . . 26 3.3 Plot of Galbumin extravasation as a function of distance from blood vessel in HCT- 116 tumour xenografts. 4 time points from injection to sacrifice are shown and the general trend is that over time, Galbumin equilibrates in the tumour. . . . . . . . 26 vii List of Figures 3.4 Plot of Hoechst extravasation as a function of distance from blood vessel in HCT- 116 tumour xenografts. 3 time points from injection to sacrifice are shown. Com- pared to the Galbumin curve (figure 3.3) Hoechst tends to extravasate further away from blood vessels (smaller molecule) . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.5 Galbumin relaxivity data plotted and fitted according to equation 1.6 and the process laid out in that chapter. We determined the relaxivity of Galbumin to be 86.62 ±5.26 (mM · sec)−1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.6 In this tumour, Gd-DTPA is the contrast agent being administered. T1-weighted gradient echo images before (A) and after (B) Gd-DTPA injection. Note the near uniform enhancement throughout the tumour - because Gd-DTPA is not a blood pool agent (diffuses everywhere), it provides MR signal enhancement everywhere. 29 3.7 A & B T1-weighted gradient echo before (A) and after (B) galbumin administra- tion. It is difficult to see an enhancement just with images (A) and (B) so (C) shows the signal enhancement parameter map from DCEMRI acquired during Galbumin administration (initial area under the curve IAUC60). Note areas of peripheral enhancement, potentially an area rich with larger vessels reside. . . . . . . . . . . 29 3.8 Apparent Diffusion Coefficient (ADC) map represents typical HCT-116 (left, [A]) and HT29 (right, [B]) tumours. Greyscales are scaled from 0-255, 0 is black and 255 is white, corresponding to an ADC ranging from 0 - 2 µmm 2 ms respectively. The HT29 tumour (right) has relatively low water difusion and equivalently, less necrosis. The tumour is outlined in red. . . . . . . . . . . . . . . . . . . . . . . . . 32 4.1 Inhomogeneities caused by the superparamagnetic contrast agent. . . . . . . . . . 33 4.2 Dual Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.3 The data acquired for the experiment to determine Galbumin susceptibility. The final value ∆χ was experimentally determined to be: 2.95 ppm/mM . . . . . . . . 35 4.4 This figure details the use of Feridex in the mouse, the parameter maps are at the top left, the equations used to construct the maps on the right and a schematic of the intensity at the bottom. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.5 The scale bar is 10 µm and the image has been taken under a light microscope by Jennifer Flexman et. al [5]. The red and purple arrows indicate different ares of staining in a neuron. Jennifer Flexman assisted with developing a protocol for Feridex staining. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 viii Acknowledgements I would like to take this opportunity to thank my co-supervisors, Dr. Stefan Reinsberg from the UBC Physics and Astronomy Department and Jennifer Baker, Ph.D candidate from the BC Cancer Research Centre. They have put in a tremendous amount of work into ensuring that this undergraduate thesis project became and stayed a reality and also supported me in presenting my research at six local, national and international scientific conferences. Without their support and commitment to helping me with the experimental work and direction, this project would likely not have happened. I would also like to thank Dr. Andrew Minchinton, a PI at the BCCRC and Jennifer Baker’s supervisor for his financial contribution in support of this project as well as invaluable feedback on practice talks, posters and proposals. Dr. Piotr Kozlowski and Andrew Yung at the high field MRI group at UBC also guided the progress of this thesis and we acknowledge their contributions to the cause. ix Dedication This thesis, the culmination of 5 years of undergraduate education, is dedicated to my parents for their perseverance in tolerating my abnormal schedules, unreasonable demands, expectations and facilitating and funding my education. Thank you. 1 Chapter 1 Introduction, Motivation and Theory 1.1 Introduction and Motivation Stopping the progress of cancerous tissue in the human body is one of the most actively researched topics in scientific research [6, 7, 8, 9, 10] . The spread of cancer has been attributed to many things and as such, there are literally thousands of branches researchers take to reach the same eventual goal - stop the inception and progress of cancer. Andrew Minchinton’s lab at the BCCRC is particularly interested in mapping tumours and studying the complex organization of their blood vessels [11]. In studying cancer and tumours, traditional biochemical methods call for analyzing frozen cross sections of tumour tissues [12, 13], staining and then fluorescently imaging them at high resolution. While this method certainly has its merits, situations and conditions are arising that require dynamic imaging of live animals. Recent advances in the field of Biophysics have allowed researchers the ability to correlate images taken with Magnetic Resonance Imaging (MRI) to those using high-resolution fluorescent microscopy. While live imaging is possible using MRI, to reproduce much of the biologically relevant data acquired by fluorescent microscopy is near impossible. In this project, we begin first by setting up the biological problem, then proceeding through and developing a protocol using the principles of MR to attempt to solve it. Figure 1.1: On the left, is a schematic of normal tissue vasculature. Notice in particular, approxi- mate symmetry and order compared to the cancerous tissue, where the vasculature is non-uniform, inefficient and wasteful (in terms of nutritive delivery). Note also in the background, the growth of ordered layers of normal cell on the left vs. the rapidly dividing cancer cells on the right. Image credit: [3] 2 1.2. Theory - Blood vessel and tumour biology The motivation behind this project is ultimately to better understand how blood vessels behave in tumours (referred to often as tumour vasculature, figure 1.1). The process where frozen tissue cross sections are analyzed [13] is called immunohistochemistry as the tissue sections are stained (see section 1.2.2) to mark for a particular molecule (growth factor or protein) a cell expresses [11, 14]. The marker is then imaged fluorescently at high resolution, and vast amounts of data can be acquired. However, this process doesn’t lend itself well for imaging dynamically to measure (for example) flow, study tumour evolution or even characterize response to combinations of anti-cancer therapies [15, 16]. With this study, our goal is to use magnetic resonance imaging to perform live scans of animals using a brand new contrast agent, Galbumin. This is an exciting development as for the first time, there is now a potential for direct pixel-pixel correlation between live MR image data and high resolution histology data. Combining these two analyses and imaging techniques is at the crux of our solution [13]. The implications for such a protocol are staggering, mainly because MR allows repeated scans for longitudinal studies and histology allows for high resolution data acquisition for cellular level resolution. Figure 1.2 shows a typical layered image from an immunohistochemical analysis. 1.2 Theory - Blood vessel and tumour biology In contrast to normal tissue, tumours are highly variable but are generally characterized by a vastly disordered organization of blood vessels and capillaries (figure 1.1) [17, 18]. Among other things, tumour vasculature is aberrant and is characterized by a deregulation of normal cell cycle processes. These abnormal processes allow us to target tumour cells specifically with radiation and chemotherapies. For example, a class of drugs could target blood vessels by inhibiting (or stimulating) angiogenesis, the creation of new blood vessels. Tumour vasculature is heteroge- nous and the uptake of compounds (glucose, oxygen, anti-cancer drugs etc.) is related, in some unknown way, to the permeability of these blood vessels. An added complication is the existence of many tumour types, each with several distinguish- ing characteristics. Here, for simplicity, we use two tumour models and exploit the intrinsic differences between them. The names of tumour types are not relevant, but tumour class A has high vasculature density (lots of blood vessels over a small area) while tumour class B has low vasculature density. This differential allows us a convenient way to test, qualitatively the effect of permeability directly. Because we want to study blood vessels in tumours, we want to use a Contrast Agent that stays inside blood vessels so we can extract information such as flow and permeability using MRI. With this study, we hope to validate and characterize Galbumin as a valid high molecular weight contrast agent and apply its unique properties as a suitable contrast agent, as well as a useful tool in tumour mapping to the two cell types. In other words, we want to use the live 3 1.2. Theory - Blood vessel and tumour biology Figure 1.2: Histology analysis - image of a tumour section where individual cells can be resolved. The colours on the tumour are explained in the legend on the right but the main idea here is that around active blood vessels (stained dark blue), there is cell growth. Cells close to the blood vessels, are more likely to receive nutrients and are therefore actively dividing. The green rim, no less than 150 µm away from blood vessels, represents cells that lack oxygen because it doesn’t diffuse that far. Above is an overlay of multiple images imaged at different fluorescent wavelengths. Data credit: Alastair Kyle (AIM lab), BCCRC imaging capabilities of MR, combined with traditional immunohistochemistry techniques to get a more accurate picture of flow in the tumour microenvironment using Galbumin, a high molecular weight contrast agent. 1.2.1 Blood Vessels, Extravasation and drug delivery It has been claimed that vessel leakiness influences the tumour microenvironment, access to therapeutic antibodies and perhaps even angiogenesis [19]. Tumour vessels are at least an order of magnitude leakier than normal vessels and there is general agreement [19, 20, 21] that these vessels are highly abnormal, however, the mechanisms involved are not very well understood. The implications of these vessels, with irregular diameters and branching patterns to cancer growth and metastasis rate has also largely been ignored, despite enormous advances in tumour angiogenesis. Water and other small water-soluble solutes see the largest variation in permeabilities among various vascular networks. In particular, factors such as the number of tight junctions in the endothelium lead to different permeabilities with the lowest values found in the brain (blood brain barrier and much higher in the intestines. It is also important to understand why size matters when studying permeability. Small solutes move through mainly by Diffusionwhereas larger molecules move through by convection and extravasation is coupled to flow and permeability. 4 1.2. Theory - Blood vessel and tumour biology Researchers[22, 23] are highlighting the importance of distinguishing between the permeability of a vessel wall vs. the “accumulation of extravasated fluid and solutes outside vessels” and [19]. Ostensibly, these accumulated fluids may be present due to a variety of reasons including but not limited to forces that drive solutes across vessel walls, and the permeability surface area product [1]. In the tumour vasculature community, there exists a school of thought that is of the opinion that all efforts should be made to “normalize the tumour vasculature” [19] before attempting to treat the tumour using traditional chemotherapy drugs. This seems slightly counter intuitive at first, however a normalization of vessels may in theory, increase blood flow by eliminating vessel abnormalities such as irregular branching and outlier diameters. One problem with this theory is that it may also stimulate tumour growth as nutrient flow is made more efficient as well. In order to test similar theories on the implications of vasculature in tumour growth, it is clear that tumour blood vessels will be at the forefront of the tumour microenvironment community. Another interesting feature of tumour vessels is that the extravasation of proteins from vessels is slower than what it would be in a normal blood vessel with pores the size of typical tumour vessels [19]. This suggests that there is an additional physical process occurring at the cellular level. It turns out that due to the high interstitial hydrostatic pressure within tumours, there is an unusually small hydrostatic gradient from the inside of vessels to the outside. This gra- dient reducing the forces that drive convection of the molecules. Ultimately, it causes proteins (macromolecules) to move out of vessels mainly by Diffusion, a slower process for larger molecules. Using the rationale above, it now is clearer why macromolecule size is a factor in studying vessel permeability. In this study, we will use a traditional MR contrast agent (Gd-DTPA) attached to an albumin protein to increase its effective size and take advantage of its low permeability across tumour vessels. 1.2.2 Immunohistochemistry Immunohistochemistry is the process that refers to localizing biomolecules in cells or frozen tissue sections. Often, specific antibodies that bind to particular antigens are used in histology procedures. The antibodies have attached fluorophores that fluoresce at specific wavelengths. In this project, histology is used to localize the MR contrast agent as well as blood vessels (using CD31 and carbocyanine / hoechst). Tumours are first excised from a Mouse Type, then frozen and sectioned using a cryostat (figure 1.3) with a thickness of 10 µm and imaged using a digital microscope. The tumours were typically 25 - 50 mm2 in area and were captured at a resolution of 1 pixel/mm2. Once the images were acquired the layers were overlaid, the CD31 and carbocyanine (or hoechst) layers were thresholded. The images show CD31 vessels that are perfused (dark blue) surrounded by regions of perfusion (light blue), unperfused vessels (red), and BrdUrd labeling (greyscale). An unperfused vessel is defined as any CD31 positive area that is negative on the 5 1.2. Theory - Blood vessel and tumour biology Figure 1.3: Image of a tumour frozen in an OCT medium being sectioned at a cryostat machine. 10µm sections adhere to glass microscope slides due to a temperature gradient. carbocyanine layer (has no light blue staining). Glowing Galbumin is a commercially available (BioPal, USA) contrast agent suitable for MR imaging. It is the low molecular weight contrast agent Gd-DTPA (Gadolinium-DiethyleneTriaminePentaacetic Acid) labelled with a high molec- ular weight bovine albumin protein (≈ 74 kDa). A schematic representation (i.e. not to scale) of Galbumin is given in figure 1.4. Glowing Galbumin, has the added advantage of 3-5 fluorescent FITC tags conjugated to Galbumin that allow for a convenient immunohistochemical analysis to localize the molecule following the study. Figure 1.4: A schematic representation of the Galbumin molecule. 15-20 Gd-DTPA atoms sur- round a large 74 kDa Albumin protein and 3-5 fluorescent FITC tags allow for convenient histo- logical imaging. 6 1.3. MR Theory 1.3 MR Theory The basic principles of nuclear magnetic resonance (NMR) involve the intrinsic spins of nuclei and their behaviour in the presence of strong external magnetic fields. Microscopically, atoms consist of both a nucleus (composed of protons and neutrons) and orbiting electrons. A net angular momentum arises in atoms that have an unmatched composi- tion of protons and neutrons (referred to as “MR-active” nuclei, e.g. 1H 19F, 23Na, 31P etc...). Semi-classically, each MR-active nucleus also has a distribution of non-zero charge (protons) and because the nucleus has an intrinsic angular momentum, a magnetic moment is induced. Consider a sample of water with an abundance of protons: at rest and unperturbed, the sample contains protons whose nuclei (figure 1.5) are oriented in all possible directions and the net angular mo- mentum of all nuclei is zero. This is because on average, for each spin orientation there exists another spin oriented in exactly the opposite direction. Similarly, the net magnetization is also zero. Figure 1.5: On the left, a typical proton is shown with its magnetization vector pointing along the z-axis. On the right, the axes are rotated so the schematic is showing the proton’s spin vector, projected along the x-y-axis. 7 1.3. MR Theory Figure 1.6: Classical diagram showing spins lining up to the external magnetic field B0 resulting in a net magnetizationM0. The spins align parallel and anti-parallel due to a small difference in energy, as shown by the energy splitting diagram (due to Zeeman splitting). However, after some time in the presence of a strong magnetic field (B0), the nuclei tend to align their axis of rotation to the applied magnetic field (figure 1.6). Quantum mechanically the picture is slightly different: the 1H is a spin-12 particle and exist simultaneously in two possible spin states, spin-up and spin-down. The spin-up state (s=−12 in figure 1.6) is more energetically favourable, since it has a slightly lower ground state energy. However, the energy difference between the two eigenstates is less than random thermal fluctuations at room temperature (E = KBT ) so slightly less than half of the nuclei are in the spin-down state, aligned anti-parallel to the external magnetic field. The small excess of nuclei in the spin-up state, are exploited in NMR. The ratio of these populations can be approximated by differences in the Boltzmann energy distributions, nup ndown = e ∆E kBT (1.1) Even at a relatively strong magnetic field of 7 Tesla, the difference in the number of nuclei is about 1 in 20,000. MR measurements are made by measuring the time it takes for perturbed magnetizations to decay back to equilibrium. However, after the magnetic field is exerted on the sample, the longitudinal component of the net magnetizationM eventually points along the direction of external magnetic field (see figure 1.5), making it almost impossible to measure. This 8 1.3. MR Theory is why MR measurements are made in the the transverse direction (x and y if the B is directed along +z). Since all of the atoms are spinning out of phase with each other, the net transverse magnetization is initially zero. Longitudinally, the initial magnetization vector (M0 can be tipped away from its thermal equilibrium value by applying an orthogonal oscillating RF pulse tuned at the Larmour frequency. The time it takes for the longitudinal component of the magnetization vector to decay back to equilibrium (aligned with B0) is referred to as the spin-lattice relaxation time or in short, T1 relaxation. Initially when the sample is first placed inside the magnetic field (before the spins align either parallel or anti-parallel to B0 oriented in the z-direction), the spins precess around B0. The spins eventually reach equilibrium and the parallel and anti-parallel orientations yield in a net magnetization vector M0. If an oscillating magnetic field is now applied to the sample in the x or y-axis, the magnetization vector is simply rotated. The duration of this oscillating magnetic field, usually in the radio-frequency range (RF - tens of Mhz) controls the extent of the magnetization vector rotation. For instance, a 90◦ pulse tips the pulse from the z-axis to the x-y-plane and a 180◦ or pi pulse tips it from +z to -z. Once rotated, due to the persistent presence of the external magnetic field, the nuclei spins immediately begin dephasing (T2 relaxation) and the magnetization in the z-direction is re-established (T1 relaxation). To measure T1 and T2 relaxation, various pulse sequences are used to measure the free induction decay (FID) in the transverse direction. Typical sequences are shown in figures 1.7 and 1.9 to measure longitudinal and transverse relaxation. 1.3.1 Measuring T1 Relaxation (using a Spin Echo Pulse Sequence) One way to measure T1 is to use a Spin Echo pulse, shown in figure 1.7. The induced voltage in the coil is measured as a function of time and then fourier transformed to given intensity. These intensities are then plotted against the “variable delay times”, and a decaying exponential is fitted to the data to produce the longitudinal relaxation time, T1. 9 1.3. MR Theory Figure 1.7: Schematic representation of a Spin-Echo pulse sequence. Multiple iterations of the spin-echo sequence (180◦ then 90◦) pulses are performed and the echo is recorded as a function of time. Figure 1.8: The signals (from figure 1.7) are Fourier transformed and here, the max signal intensity (of each echo) is plotted as a function of the variable delay. The time constant from the exponential recovery is T1 . More precisely, T1 is measured as the time it takes for the signal intensity to return to 63 % of its’ maximum value. 10 1.3. MR Theory 1.3.2 Measuring T ∗2 Relaxation using (CPMG Pulse Sequence) T ∗2 is commonly measured using a pulse sequence called a CPMG sequence. Figures 1.9 and 1.10 detail the sequence and the method for determining T ∗2 and T2 using the CPMG sequence. Notice that T ∗2 is always shorter than T2 because the T ∗2 relaxation time includes spins dephasing due to both spin-spin interactions as well as field inhomogeneities. Figure 1.9: Schematic representation of a CPMG pulse sequence. The initial 90◦ pulse tips the magnetization pulse in the transverse direction, the magnetization then decays rapidly due to field inhomogeneities and other spin-spin interactions. Application of a 180◦ pulse flips the direction of the dephasing spins and causes them to rephase. As the spins rephase, a spin echo signal is generated. The T2 decay curve in blue is represented in figure 1.10. 11 1.3. MR Theory Figure 1.10: The curve above shows the T2 relaxation process. The FID signal from figure 1.9 is fourier transformed and the maximum intensity is plotted as a function of the Echo Time (TE). The decay constant is the spin-spin relaxation rate. As shown in figure 1.11, when a magnetic moment is tipped so the angle between the mag- netization vector and the original magnetic field is α, the field exerts a torque on the vector M causing it to precess. This precession frequency is referred to as the Larmor frequency and is given by, ω0 = γB (1.2) Here, γ is the gyro-magnetic ratio, an intrinsic property of the nucleus and B is the magnetic field, dominated by the static field B0 with a small contribution from a magnetic field gradient. This magnetic field gradient is the key to NMR imaging as it changes the total magnetic field strength at a particular spatial location and thus, by equation 1.2, the Larmour frequencies of the nuclei vary by location and the spatial coordinates can be encoded into the measured resonant frequencies. The following general description show some basic NMR procedures to image a sample. Basic Steps for acquiring MR data 1. Place sample in constant magnetic field B0, length aligned parallel to B0 and the z-axis (figure 1.11). 2. Apply a continuous magnetic field gradient across the sample 3. Simultaneously, apply a radio frequency (RF) pulse close to the larmour frequency ω0. 12 1.3. MR Theory Figure 1.11: On the left, the magnetization vector tipped and as it returns back to B0, it precesses around B0. [4]. On the right, a schematic of slice-selection with a patient lying along the a magnetic field along their length. 4. The magnetic field gradient causes each slice along the z-axis to have a slightly different net B and by equation 1.2, a different Larmour frequency. 5. Recall that the RF pulse tips the magnetization vector so with the RF pulse at ω0, only a particular slice has its’ vector M tipped and all the other nuclei are still aligned with the B field they encounter (function of distance along z). 6. By tuning ω to match ω0, we effectively “select” the slice we’re interested in imaging. 7. The magnetization vector precesses around B0 and eventually decays back to its steady state (aligned with B0). The longitudinal component of this relaxation is referred to as T1 and the transverse component as T2. 8. Receiver coils in the MR scanner pick up the voltages induced by the precessing magneti- zation 9. The FID’s (or echoes with T1) are fourier transformed and the maximum intensities are extracted and plotted as in figure 1.9. 10. Similar gradient and pulse sequences in x,y and z can be used to encode frequency and phase so the final image looks like figure 1.12. The human body is composed of about 60% water, so hydrogen nuclei are a natural choice for MR imaging experiments in both humans and animals. Contrast agents allow a further enhancement and with T1 or T2 weighted imaging, can increase the amount of detail present in images taken in-vivo. In this study, we will be considering two types of weighted imaging, and our novel protocol will require the use of both T1 and T ∗2 weighted imaging simultaneously. 13 1.4. MR Contrast Agents Figure 1.12: Left: representation of the FID signal data acquired, notice concentration of data around the origin. Right : Typical MR image of a tumour after a Fourier Transform. Bright bulb at the bottom is the crosssection of a vial of water, used as a standard reference point. 1.4 MR Contrast Agents MR images are created based on the signal from MR-active nuclei relaxing under the presence of an external magnetic field. Image contrast is primarily based on the distribution of these nuclei inherent in different tissues (e.g. water or lipid content). To increase contrast, MR contrast agents are used to selectively alter the relaxation times of hydrogen nuclei in tissues where the contrast agent is present. The driving force behind using MR contrast agents (and other imaging modalities) is to improve image contrast and visualize areas that would otherwise not appear on the scan. This can often occur when the tissue properties of an area of interest is too similar to that of its surroundings and the effect averaged out. In a T1 weighted image, areas with short T1 relaxation times such as fat, appear bright and water with significantly higher T1 relaxation times, appear dark [24]. In other imaging modalities, such as X-ray or CT scans, iodine or barium contrast agents are used that result in a direct effect in image contrast (tissues containing barium or iodine results in greater attenuation of x-rays. MR contrast agents work indirectly as no signal is derived from the contrast agent itself - the relaxation times in hydrogen nuclei is still measured but the times have been altered due to the proximity of the nuclei to the contrast agent. The best MR contrast agents therefore are those that produce the largest gradient in relaxation times. Table 1.4 categorizes contrast agents into three major groups: Recall that the spin-lattice (T1) and the spin-spin (T2) relaxation times are related to their relaxivities (relaxation rates) reciprocally: 14 1.5. Modelling Contrast Agents - Kety Model CA Type Effect on T1 Effect on T2 or T ∗2 Examples Paramagnetic Agents decreases decreases Gd-DTPA Superparamagnetic particles None Significant Decrease Feridex, Combidex Ferromagnetic particles None Dramatic decrease fluidMAG Table 1.1: A table of the typical divisions of MR contrast agents [2]. In this study, we use Feridex, Gd-DTPA and Galbumin (mainly a T1 agent). Note, the table doesn’t quantify the difference between “significant” and “dramatic”, because there is a range of effects depending on concentration and relaxivities. r1 = 1 T1 and r2 = 1 T2 (1.3) The relationship between relaxation rates and contrast agent concentration can then be pre- dicted by the Solomon-Bloembergen equations [22, 25]: 1 T1 = 1 T10 + r1 [Gd] (1.4) 1 T2 = 1 T20 + r2 [Gd] (1.5) where [Gd] is the contrast agent concentration, usually a Gd-based agent. Using equation 1.5, we can now find a relationship for CGd(t) empirically: CGd(t) = R10 −R1(t) r1 (1.6) Once the contrast agent relaxivity has been determined, assumed concentrations of CA in the blood will result in predicted T1 values. Each T1 value will yield a signal intensity (from a short-TR gradient echo sequence), given by the following well established equation [25]: S.I. = g · ρ ( 1− e TR T1 ) 1− cos(α)e −TR T∗1 · e −TE T∗2 sin(α) (1.7) 1.5 Modelling Contrast Agents - Kety Model The 2-compartment pharmacokinetic model called the Kety-model later expanded by Tofts et. al. [1] is the industry standard in modelling contrast agent uptake in humans and animals. The model was initially intended for use in the brain, where the effect blood brain barrier was a key assumption. Figure 1.13 shows a schematic of the model - the Kety model relates the movement 15 1.5. Modelling Contrast Agents - Kety Model of contrast agent between two compartments, the blood vessel (containing plasma) and the inter- stitial space. The red arrow in figure 1.13 termed Ktrans is the forward rate, and it encompasses both blood flow and (see [26] for details) the vessel permeability surface area product (PS - see [1]). The reverse rate, Kep is the rate constant between the EES and blood plasma (see Table 1.2 for details on these quantities). This model can be applied in two limiting circumstances: flow-limited and permeability lim- ited. Low molecular weight agents such as Gd-DTPA, considered freely diffusing agents transfer between the blood and the EES of the tumour with a ktrans that is proportional to the the flow rate. Larger, minimally diffusing agents result in a Ktrans that is proportional to the permeability surface area product PS. Parameter Units Physical Meaning EES none Extravascular extracellular space ve none Volume of EES per unit volume of tissue Ktrans min−1 Volume transfer constant Kep min−1 Rate constant between EES and blood plasma Cp mM Tracer concentration in arterial blood plasma Ct mM Tracer concentration in tissue Table 1.2: EES refers to the space into which tracer can leak from a capillary and Ktrans is the volume transfer constant between blood plasma and the EES. Table parameters by: [1] Figure 1.13: Two compartment Kety Model with rate constants describing flow of contrast agent out into the EES (compartment A)and back into the vessels (compartment B). 16 1.5. Modelling Contrast Agents - Kety Model The kinetic tracer model can best be described in the following form [1]: dCt dt = Ktrans(Cp − kepCt) (1.8) When Ct = Cp = 0 at time t=0, equation 1.8 solves to the following relation: Ct(t) = Ktrans ∫ t 0 Cp(τ)e−kep(t−τ)dτ (1.9) The dual contrast agent protocol we hope to implement has this idea at its’ core - in order to determine the permeability of tumour vessels, two agents may be used in succession. The first, a small molecular weight agent would yield a Ktrans related to both permeability and blood flow. Then, in the same mouse, a second high molecular weight agent would yield in a Ktrans - ostensibly related only to the blood flow as the agent is too large to extravasate out of vessels. Figure 1.14 shows a schematic of this protocol. Figure 1.14: Schematic of a dual contrast agent protocol - the first contrast agent is injected, the signal acquired and then the second agent is administered. Because both sets of data have been acquired, the signal intensities can be processed and the two processes can be treated as independent. 17 Chapter 2 Materials and Methods In summary, 5 NOD/SCID mice with subcutaneous HCT-116 (3 mice with human colorectal cancer) and HT29 (2 mice with human Caucasian colon adenocarcinoma) xenografts received an initial scan (under anesthesia) for water-diffusion. The mice then received a baseline T1 weighted MRI scan prior to injection of the first contrast agent. Then a dynamic scan was taken for 30 minutes with the low molecular weight agent (Gd-DTPA). A follow up T1 weighted scan was taken to re-establish a new baseline T1 and were then injected with the second contrast agent and another dynamic scan was taken. After the scans, the mouse was removed from the scanner and hoechst was injected i.v. 5-20 minutes before excision. 2.1 Mouse Protocol The following is a reasonably accurate procedure for implanting the mice with tumours (for more details see [8, 11, 27]: 1. Seed 2 x T-175 flasks with 1.5 x 106 cells (1:10) days before implant 2. Split to 4 x T-175 flasks in 3-4 days or whenever 70% confluency is achieved 3. Collect cells to 1.6 x 108 cells / mL (for 8 x 106 cells per 50µl implant) 4. Anaesthetize mice 5. Sterilize sacral region using isopropyl alcohol 6. Implant 50 µl cells per mouse 7. Return mice to cage 8. Measure tumours once or twice a week 9. Tumours are ready in 3-4 weeks 2.1.1 Preparing a mouse for an MR scan Animal handling was performed according to the ethics guidelines approved by the Animal Care Committee valid until December 31, 2008. The animals were stored at the Animal Research 18 2.1. Mouse Protocol Centre at the BC Cancer Research Centre and transported to the 7T MRI facility at the UBC Life Sciences Centre the day of the experiment and were sacrificed at the end of the experiment that day. Refer to figure 2.1 and the following outline for the animal preparation procedure. 1. Begin Mouse Prep Protocol 2. Turn on water blanket to thermoregulate anaesthetized mice while in the MR scanner (45◦C set point, pump = fast). 3. Turn on ParaVision (imaging software), electronics, unlock magnet room, check anesthetic levels, refill anaesthetic if necessary, turn on medical air. 4. Record information on mouse: Type (HCT116 or HT29), number of tumours (1 or 2), age, gender, identification code, tail markings. 5. Weigh mouse (and record weight) prior to scan and injections 6. Insert catheter into tail vein with appropriate connector. 7. Use heat lamp to warm up mouse and increase blood flow (helps identify tail vein for injection) 8. Prepare solution of heparin (anti-coagulant) diluted in saline 1:10 9. Inject 200-250 µL of saline into mouse to keep it hydrated 10. Place mouse into anaesthesia box, wait about 30 seconds 11. Squeeze toes to check if suitably anaesthetized 12. Check breathing 13. Remove mouse from restraint, make sure it is still anaesthetized. Leave on heat lamp. 14. Apply protective lubricant on eyes to prevent drying/scratching. 15. Move mouse from the anaesthesia box to the scanning coil 16. Place mouse’s mouth and nose over the anaesthetic hood on the coil (make sure mouthpiece is in the proper orientation. 17. Adjust mouse position to have the tumour(s) in place directly inside the coil 18. Insert rectal temperature probe, use Vaseline on the end of the probe. Tape down. 19. Transport mouse from the animal room to the scanning room, attach anaesthetizing tube in the scanning room to the connector on the coil 19 2.2. Histology Figure 2.1: Description of Experimental setup 20. Attach respiratory monitor onto mouse with tape. and tape the catheter, syringes any other wires/tubes to the bed. 21. Place water blanket over mouse, check temperature and breathing rate 22. Monitor respiratory rate and temperature. Monitor anesthetic levels (≈ 1 unit of air, 0.5 to 1 unit of isofluorane). 23. End Mouse Prep Protocol 2.2 Histology An automated microscope with x-y-z slide stage was used to acquire entire tumour-section images. Hoechst (perfusion), CD31 (vasculature), Haematoxylin (nuclei) and Galbumin (contrast agent) were fluorescently imaged independently from tissue sections corresponding to MR scan slices (see [11, 27] for details). Galbumin and Hoechst distribution was mapped as a function of distance away from blood vessels in the two tumour types. Extravascular extracellular space, marked with Coll IV was measured and correlated with ADC from diffusion-weighted EPI. 20 2.3. MR Scans 2.3 MR Scans MR scans were taken at the high field MR research centre at the Life Sciences Institute under the direction of Dr. Stefan Reinsberg, Dr. Piotr Kozlowski and Andrew Yung. Imaging was per- formed on a 7T Bruker Biospec 70/30 using a custom-built 4 turn distributed-capacitor solenoid. The final dual contrast agent MR protocol is summarized below: 1. Tripilot (scout) (2 minutes) 2. Diffusion weighted EPI (3-4 minutes) 3. High-Res image for baseline T1 (5 minutes) 4. Begin acquisition: T1 DCE-MRI (30 minutes) 5. Administer contrast agent 1 after at least 30 seconds of acquisition 6. High-Res image for T1 (5 minutes) 7. Begin acquisition: T1 weighted DCE-MRI (30 minutes) OR T∗2 weighted DSC-MRI (30 minutes) 8. Administer contrast agent 2 after at least 30 seconds of acquisition 9. High-Res image post-contrast T1 (5 minutes) 2.3.1 Diffusion-weighted EPI Diffusion-weighted EPI with TR/TE= 3000/26.9, b-values 0,500 was performed and apparent diffusion coefficient (ADC) maps were calculated. See figure 3.8 for a representative set. 2.3.2 T1 weighted images 30 minutes of the T1-weighted FLASH (TR/TE=113/2.145) FLASH were acquired following a 100-150µL/mouse bolus of Galbumin (25 mg/mL). A two-TR FLASH protocol (TR=226 ms and 113 ms, TE= 2.145 ms) was used for calculating contrast agent concentration with time resolution of 14.5s. Scanning was conducted for 30 minutes because unlike with T ∗2 -weighted imaging, enhancement lingers as the agent equilibrates over the tumour. Because the the effect of the T ∗2 agent relies upon introducing magnetic field inhomogeneities, the effect rapidly decreases (signal intensity recovers quickly) as it circulates through the blood stream. 21 2.4. Galbumin Relaxivity 2.3.3 T∗2 weighted images 150s of the T ∗2 -weighted FLASH (TR/TE=46.875/10.0) FLASH were acquired following a 100- 150µL/mouse bolus of Galbumin (25 mg/mL). Flip angle was set to 30◦ and two slices of the tumour were acquired with a time resolution of 150s for complete acquisition. 2.4 Galbumin Relaxivity Referring to equations 1.5 and 1.6, we can construct a protocol to experimentally determine the relaxivity of Galbumin if we take several samples with a known concentrations of Galbumin, record measured T1 values and plot the signal intensity (equation 1.7). The protocol used is described below: • Prepare vials with at least 5 precisely known concentrations of contrast agent • Perform a T1 weighted MR scan on each of the vials independently • Record T1 values for each of the 5 known concentrations • Plot R1 vs. CGd(t) and fit to a linear function (f(x)=mx+b) • Slope of this graph (figure 3.5) is the relaxivity 22 Chapter 3 Results, Analysis and Discussion 3.1 Galbumin Histology Galbumin (commercially available from BioPal, U.S.A) is a contrast agent that lends itself well to histological analysis immediately after the tumour has been frozen. In other words, once the MR scan is complete (≈ 45 minutes) and the mouse has been sacrificed with the tumour excised and frozen (≈10 minutes), the tumour can be sectioned and directly imaged (fluorescently) for Galbumin. This is the most direct method of imaging as it avoids several washing and fixing steps that may potentially result in washout artefacts for the target molecule (Galbumin in this case). To localize Galbumin with respect to other histological features such as blood vessels however, the washing and fixing steps are required. Due to the versatility of the FITC tags, the slides can be imaged again to account for intensity differences, bleaching and washout effects. Figure 3.1 shows a typical tumour section stained for blood vessels (red, CD31) and Galbumin (green, FITC). In figure 3.3, we mapped the presence of Galbumin (detected by positive fluorescence intensity) as a function of distance from the nearest perfused vessels(determined by the presence of both CD31 and hoechst). Contrast figure 3.3 - mapping a large molecule to perfused vessels, to figure 3.4, where the presence of hoechst is mapped as a function of distance from perfused vessel in the same tumour. Qualitatively, most of the Galbumin is localized to within 50µm away from vessels whereas Hoechst, the smaller molecule, perfuses/diffuses as far as 100 µm away from vessels. Quantitatively, there exist several problems with comparing extravasation curves. Primarily, there exist intrinsic differences in the molecules’ ability to fluoresce at particular wavelengths. It may be that the tissue exposed to 488nm light, may have a higher background compared to 546nm or 350 nm. Further, the exposure time of tissue to the fluorescent lamp greatly influences the fluorescence signal Intensity. Finally, to accurately compare two extravsation curves and extract meaningful quantitative information, we require there to be several controls imaged at various exposure times, with and without the target molecule to predict the contribution of background fluorescence. The tumour microenvironment is heterogeneous and the vasculature disordered and in fig- ures 3.1and 3.2 there is a clear differential in vasculature behaviour among vessels (in the same tumour). Since the differential appears to be uncorrelated with any known effect, this suggests that chemotherapeutic drugs administered would likely reach their target at different times, if at all. This idea is of crucial importance because in order for a tumour to be fully eradicated, 23 3.1. Galbumin Histology all cancer cells must be killed. One can imagine that if drugs take effect in certain areas earlier, surrounding cells may compensate for this and the effect of the drug would effectively, be reduced drastically. A possible solution, and one that is used often, is the drug is administered in doses much larger than strictly required to ensure maximal effect. Of course, this has the disadvantage of increased toxicity and other unintended physiological consequences. Future studies would look at the difference in vessels in various areas and attempt to charac- terize their behaviour more accurately. The ideal method for accomplishing this is a long term (longitudinal) study on a group of mice that tracks and characterizes the behaviour of the tumour and vessels over time. Pike et. al.[? ] conducted a similar study and looked at tumour growth over time. We intend to extend that study to investigate possible explanations to vasculature differential including the age of vessels, proximity to other established/unestablished vessels as well as a fundamental characteristic of the tumour cell line. We used mainly HCT116 (human colon carcinoma) and HT29 (human colon adenocarcinoma grade II). Comparing vasculature be- haviour more rigourously among these two cell lines using histology and MRI is a potential goal for this study. 24 3.1. Galbumin Histology Figure 3.1: Typical tumour (HCT116) section stained for blood vessels (red, CD31) and Galbumin (green, FITC). Sections of the same tumour are stained and (fluorescently) imaged separately and overlaid on top of each other. 25 3.1. Galbumin Histology Figure 3.2: An area of the tumour above (figure 3.1) zoomed in. We note some vessels (red, CD31- bottom left) have positive Galbumin staining (green, FITC) while other vessels show no positive staining (top right). Figure 3.3: Plot of Galbumin extravasation as a function of distance from blood vessel in HCT- 116 tumour xenografts. 4 time points from injection to sacrifice are shown and the general trend is that over time, Galbumin equilibrates in the tumour. 26 3.2. Galbumin as T1 agent Figure 3.4: Plot of Hoechst extravasation as a function of distance from blood vessel in HCT- 116 tumour xenografts. 3 time points from injection to sacrifice are shown. Compared to the Galbumin curve (figure 3.3) Hoechst tends to extravasate further away from blood vessels (smaller molecule) 3.2 Galbumin as T1 agent The initial intent of the dual contrast agent injection protocol was to use Galbumin as the high molecular weight agent and Gd-DTPA as the low molecular weight agent. Because Galbumin is a new product on the market, it had to be characterized in terms of its effect on T1 relaxation. In figure 3.5, we determined the relaxivity of Galbumin to be 86.62 ±5.26 (mM · sec)−1, much higher than published values for each Gd-DTPA molecule at 8.5T (we used a 7T magnet) 3.87 0.06 (mM · sec)−1 [28]. However, this differential is easily accounted for when considering the composition of the Galbumin molecule. BioPal claims that anywhere from 15-20 Gd-DTPA atoms are attached to an Albumin protein. The adjusted value for Galbumin per Gd-DTPA molecule results in a value ranging from 4.33 to 5.77 (mM · sec)−1, closer in line with published values. 27 3.2. Galbumin as T1 agent Figure 3.5: Galbumin relaxivity data plotted and fitted according to equation 1.6 and the process laid out in that chapter. We determined the relaxivity of Galbumin to be 86.62 ±5.26 (mM · sec)−1. 28 3.2. Galbumin as T1 agent F ig ur e 3. 6: In th is tu m ou r, G d- D T PA is th e co nt ra st ag en t be in g ad m in is te re d. T 1 -w ei gh te d gr ad ie nt ec ho im ag es be fo re (A ) an d af te r (B ) G d- D T PA in je ct io n. N ot e th e ne ar un ifo rm en ha nc em en t th ro ug ho ut th e tu m ou r - be ca us e G d- D T PA is no t a bl oo d po ol ag en t (d iff us es ev er yw he re ), it pr ov id es M R si gn al en ha nc em en t ev er yw he re . F ig ur e 3. 7: A & B T 1 -w ei gh te d gr ad ie nt ec ho be fo re (A ) an d af te r (B ) ga lb um in ad m in is tr at io n. It is di ffi cu lt to se e an en ha nc em en t ju st w it h im ag es (A ) an d (B ) so (C ) sh ow s th e si gn al en ha nc em en t pa ra m et er m ap fr om D C E M R I ac qu ir ed du ri ng G al bu m in ad m in is tr at io n (i ni ti al ar ea un de r th e cu rv e IA U C 60 ). N ot e ar ea s of pe ri ph er al en ha nc em en t, po te nt ia lly an ar ea ri ch w it h la rg er ve ss el s re si de . 29 3.3. ADC (Apparent Diffusion Constant) Maps Figures 3.6 and 3.7 show images from typical DCE (dynamic contrast enhanced) scans with Gd-DTPA and Galbumin. Despite the fact that Galbumin had a similar relaxivity to Gd-DTPA, Galbumin did not provide a large enough enhancement. Using the signal intensity equation discussed in chapter 3 (equation 1.7) we can calculate the net change in intensity from a measured change in T1 before and after Galbumin administration. With a 100 µL dose of Galbumin injected I.V., the calculated T1 value is about 418 ms. The T1 of blood plasma is around 1200 ms and with the other parameters set as in chapter 2, applying the signal intensity equation twice yields a net signal intensity increase of 151%. However, signal enhancement only occurs in vessels and areas with blood. We can estimate that about 5% of the total volume of a typical mouse at 20 g is blood volume and extending that to each voxel, only a total signal intensity increase of 7.5% is achieved. Several attempts were made to continue using Galbumin as the high molecular weight agent in the dual contrast agent protocol because it is an ideal compound to use histologically. The first strategy was to increase the dosage from 100 µL to 200 µL. This had several side-effects as the tolerance of the mouse (to the contrast agent) seemed to decrease with the added volume injected I.V. through the tail vein. More importantly, as expected the increase in intensity was statistically insignificant. Attempts to concentrate the mixture were quickly derailed because the manufacturer suggested that it would become too viscous past 4x concentrated. The already costly agent would increase in price by a factor of at least 4. Consider all this, we concluded that Galbumin was not a suitable contrast agent for T1 weighted imaging at the concentration the manufacturer provides in our dual-contrast agent protocol. One insight we’ve gained from this pilot study is that having a larger contrast agent results in requiring more of the agent to achieve the same effect as a smaller agent if the active molecules (Gd-DTPA) are the same. In searching for more suitable contrast agents, one restriction we can now place is that the relaxivity of the agent has to be several times higher. 3.3 ADC (Apparent Diffusion Constant) Maps While the mouse is in the scanner, before contrast agent administration, it is often convenient to do a simple diffusion weighted MR scan. This serves several purposes and in a longitudinal study, diffusion-weighted MR scans can be time coursed and the changes studied over time. The change in the apparent diffusion coefficient (ADC) of water is hypothesized to be an indicator of tumour necrosis. DW scans were acquired using echo-planar imaging (EPI - TR/TE=3000/26.9) and apparent diffusion coefficient (ADC) maps were calculated. Figure 3.8 shows the ADC maps for the two tumour models used in this study. HCT116 (left) is characterized by higher propensity to necrosis and leakier vessels. DCEMRI parameter maps (figure 3.7) show this higher leakiness even for macromolecular contrast agent Galbumin. HT29 (right) is characterized by moderate necrosis (corresponding lower ADC values) and reduced leakiness. In future studies, we hope to 30 3.3. ADC (Apparent Diffusion Constant) Maps make use of ADC maps in studying tumour growth and as such, included them in the protocol development stage. 31 3.3. ADC (Apparent Diffusion Constant) Maps F ig ur e 3. 8: A pp ar en t D iff us io n C oe ffi ci en t (A D C ) m ap re pr es en ts ty pi ca l H C T -1 16 (l ef t, [A ]) an d H T 29 (r ig ht , [B ]) tu m ou rs . G re ys ca le s ar e sc al ed fr om 0- 25 5, 0 is bl ac k an d 25 5 is w hi te ,c or re sp on di ng to an A D C ra ng in g fr om 0 - 2 µ m m 2 m s re sp ec ti ve ly . T he H T 29 tu m ou r (r ig ht ) ha s re la ti ve ly lo w w at er di fu si on an d eq ui va le nt ly ,l es s ne cr os is . T he tu m ou r is ou tl in ed in re d. 32 Chapter 4 T ∗2 Weighted Imaging: Galbumin & Feridex Magnetic susceptibility is a fundamental property of matter and is defined as the ability of the external magnetic field to affect the nucleus of an atom and magnetize it. The role of contrast agents in general, is to introduce changes in the local magnetic field. Using molecules with high susceptibility is one way to induce small local inhomogeneities in B and from equation 1.2, a change in the larmour (precession) frequency is caused to the nuclei(figure 4.1). The signals from the sum of these precessions collectively decay quicker than if there was no B field inhomogeneity. This is the underlying principle of T∗2 weighted imaging and referring to figure 1.2, we notice that Feridex is commonly used as an agent in T ∗2 weighted imaging (it is a superparamagnetic iron oxide particle - SPIO). However, Galbumin has not yet been eliminated as an agent affecting T ∗2 . Figure 4.1: Inhomogeneities caused by the superparamagnetic contrast agent. Our dual-contrast agent protocol can be readily adapted to T ∗2 weighted imaging (as opposed to T1 weighted, figure 1.14). Figure 4.2 shows the modification, following the T1 agent injection, we simply change the scan type and proceed as usual. The parameters measured in T ∗2 weighted imaging are slightly different so we won’t get a ktrans, but rather the rBF, rBV and MTT (see section 4.2). 33 4.1. Galbumin as a T ∗2 agent Figure 4.2: Dual Protocol 4.1 Galbumin as a T ∗2 agent The magnetic susceptibility of Galbumin was measured according to previously published meth- ods [29] and it was found that Galbumin susceptibility was not high enough to allow effective T ∗2 weighted imaging. Reference susceptibility values are shown in Table 4.1. Equation 4.4 was used to find the susceptibility ∆χ, ∆χ = − 2∆φ B0TEγproton 1( cos2(θ)− 13 ) (4.1) (4.2) ∆χgalbumin = 2.9465x10 −6 (4.3) (4.4) The parameters used in determining the susceptibility curve (figure 4.3) are below: • θ = 10.1 degs • φ = -0.60 rads • B0 = 7 T • TE = 2.145 ms = 0.002145 secs • Gyromagnetic ratio of protons γproton=42.576MHz/T, 34 4.2. Feridex as a T ∗2 agent Figure 4.3: The data acquired for the experiment to determine Galbumin susceptibility. The final value ∆χ was experimentally determined to be: 2.95 ppm/mM Substance Susceptibility ∆χ (ppm) Blood -0.3 to 1.5 Gd-DTPA 2.7 ± 0.1 AMI-25 40 ± 2 Feridex TBA Table 4.1: Table parameters from: [2] 4.2 Feridex as a T ∗2 agent In T ∗2 weighted imaging, the contrast agent (Feridex here) acts as a tracer (that stays in vessels) during at least the first pass through the circulatory system [30]. Applying the appropriate models, we can obtain the relative blood flow (rBF) and relative blood volume (rBV) within a tumour. There is also a third useful parameter, the mean transit time (MTT) which is the ratio of the rBV and the rBF. In this study, we have only acquired preliminary data using Feridex, and analysed it sparingly. 35 4.2. Feridex as a T ∗2 agent Recovery and intensity parameters are shown from two slices of an MR scan are shown in figure 4.4. Since T ∗2 weighted imaging results in an intensity dip due to magnetic field inhomo- geneities caused by the contrast agent (Feridex), we would expect to see a recovery in signal intensity once the agent has equilibrated in the blood. The recovery maps show the relationship between two intensities: middle of the scan (lowest intensity) and at the end (return to intensity) and both slices (slice 1 and 2) show no recovery in the intensity 140 seconds after injection.The intensity maps show us which region of the tumour had the largest decrease in signal intensity. The areas that are black in the parameter maps show no decrease in intensity, and consequently, we can expect that those areas saw no contrast agent flow (necrosis or interstitial space). Figure 4.4: This figure details the use of Feridex in the mouse, the parameter maps are at the top left, the equations used to construct the maps on the right and a schematic of the intensity at the bottom. 36 4.3. Mapping Feridex Histologically 4.3 Mapping Feridex Histologically Feridex (generically ferrumoxide), is a sterile aqueous colloid of superparamagnetic iron oxide (SPIO) associated with dextran. Administered intravenously, it is an MR contrast agent usually used for the detection of liver lesions. Feridex is taken up by macrophages, found only in healthy liver cells but not in most tumors. We want to use Feridex in the DSC weighted portion of our protocol but the major disadvantage of Feridex is that it does not lend itself well to being characterized histologically. As a result, the tracking and/or mapping the movement of Feridex as it extravasates out of tumour vessels. The absence of the FITC tags requires using more rigourous, less versatile staining methods (such as chemical staining). We have recently discovered that the Feridex contains a dextran coat, which could potentially be useful in manually attaching fluorescent tags or even using antibodies specific to that dextran for immunohistology. 4.3.1 Chemical Stain - Prussian Blue Prussian Blue is a synthetic pigment commonly used in histology to detect the presence of non- hemoglobin iron. More specifically, ferric ions (Fe+3) ions in the tissue combine with ferrocyanide (reagent) and results in the formation of a bright blue pigment called Prussian Blue. See figure 4.5 for sample staining [6]. The general protocol is described in the appendix A.4, Figure 4.5: The scale bar is 10 µm and the image has been taken under a light microscope by Jennifer Flexman et. al [5]. The red and purple arrows indicate different ares of staining in a neuron. Jennifer Flexman assisted with developing a protocol for Feridex staining. 37 Chapter 5 Conclusions, Implications and Future Work 5.1 Conclusions In this study, we have managed to establish, at least theoretically, a dual contrast agent pro- tocol that will serve to accurately extract tumour vasculature permeability information that is generally coupled to vasculature flow in Ktrans. Following this, we considered Galbumin, a novel contrast agent (commercially available from BioPal USA) as a candidate for our dual CA proto- col. Through both experimental and theoretical means, we established that while the relaxivity of Galbumin (4.33 to 5.77 (mM · sec)−1 per Gd-DTPA chelate) was close to the relaxivity of Gd-DTPA (3.87 ± 0.06 (mM · sec)−1 [28]), the required level of signal enhancement was not seen when imaging with the scans T1 weighted. Next, we adapted our protocol to include T ∗2 weighted imaging and thus considered Galbumin again as a T ∗2 contrast agent. Again, through experimen- tation and as expected, we found that Galbumin susceptibility (2.95 ppm/mM) was very close to Gd-DTPA susceptibility (2.7 ppm/mM ±0.1), effectively eliminating it from consideration as a candidate for a T ∗2 agent. Ultimately, Feridex, a well-studied SPIO (superparamagnetic iron oxide particle), emerged as the leading candidate for a high molecular weight contrast agent in our protocol. Preliminary experiments were conducted with Feridex in mice and it was found that the particle did not extravasate out for at least the first 150 seconds (figure 4.4). Further, it’s susceptibility was quite high and thus made it ideal for a T ∗2 contrast agent. In addition to the MR work above, we also developed, refined and optimized several histology protocols for use in the future. For example, we now have a set protocol that dictates precisely the steps to take from when the MR scan is complete up until tumour sections have been obtained. Several challenges were encountered in maintaining the tumours in the orientation they were excised to match with MR images as closely as possible. Once we have established the right high molecular weight contrast agent to use in the MR protocol and determined the most effective way of visualizing it histologically (whether it be a chemical stain/dye or a fluorescent marker), we are confident the histology portion of the study has been reasonably optimized. 38 5.2. Implications and Future Work 5.2 Implications and Future Work The work presented in this thesis is likely going to be the basis of a Masters project. A lot of the ideas presented here seemed deceptively simple at first, several challenges and roadblocks were encountered in the process of arriving at some of the conclusions. The following is a list of items that are at the top of our list to attempt over the next few months: • Study (rigourously) Galbumin extravasation from vessels • Look into using dextran-DTPA contrast agents (as high as 500 kDa) • Consider other high molecular weight T1 agents suitable for dual-injection protocol • Experimentally determine Feridex Susceptibility • Investigate labelling Feridex with FITC tags • Ensure Feridex is an intravascular agent • Investigate mapping Feridex using antibodies specific to the dextran coating • Devise and test a method to extract vessel permeability from the dual-injection protocol • Implement the use of fiducial markers in more accurately determining the orientations of tumour from MR scan to histological section • Investigate vascular permeability over time in a longitudinal study • Study qualitatively and quantitatively the effect of anti-angiogenesis drugs on vessel per- meability 39 Bibliography [1] P Tofts, G Brix, D Buckley, and J Evelhoch. Estimating kinetic parameters from dynamic contrast-enhanced t-weighted mri of a diffusable tracer: . . . . JOURNAL OF MAGNETIC RESONANCE IMAGING, Jan 1999. [2] R Edward Hendrick and E Mark Haacke. Basic physics of mr contrast agents and maximiza- tion of image contrast. JOURNAL OF MAGNETIC RESONANCE IMAGING, 3(1):137– 148. [3] Stefan Reinsberg. Reinsberg.nserc.2008. page 6, Nov 2008. [4] Qing-San Xiang. Ubc p404 class notes. pages 1–13, Sep 2007. [5] Sosuke Miyoshi, Jennifer A Flexman, Donna J Cross, Kenneth R Maravilla, Yongmin Kim, Yoshimi Anzai, Junko Oshima, and Satoshi Minoshima. Transfection of neuroprogenitor cells with iron nanoparticles for magnetic resonance imaging tracking: Cell viability, differ- entiation, and intracellular localization. Mol Imaging Biol, 7(4):286–295, Jul 2005. [6] Brad P Barnett, Aravind Arepally, Parag V Karmarkar, Di Qian, Wesley D Gilson, Piotr Walczak, Valerie Howland, Leo Lawler, Cal Lauzon, Matthias Stuber, Dara L Kraitchman, and Jeff W M Bulte. Magnetic resonance–guided, real-time targeted delivery and imaging of magnetocapsules immunoprotecting pancreatic islet cells. Nat. Med., 13(8):986–991, Sep 2007. [7] Moira C Ferrier, Moira C Ferrier, Hemant Sarin, Hemant Sarin, Steve H Fung, Steve H Fung, Bawarjan Schatlo, Bawarjan Schatlo, Ryszard M Pluta, Ryszard M Pluta, Sandeep N Gupta, Sandeep N Gupta, Peter L Choyke, Peter L Choyke, Edward H Oldfield, Edward H Oldfield, David Thomasson, David Thomasson, John A Butman, and John A Butman. Validation of dynamic contrast-enhanced magnetic resonance imaging-derived vascular permeability mea- surements using quantitative autoradiography in the rg2 rat brain tumor model. Neoplasia (New York, N.Y.), 9(7):546, Jul 2007. [8] Alastair Kyle, Lynsey Huxham, Devon Yeoman, and Andrew Minchinton. Limited tissue penetration of taxanes: A mechanism for resistance in solid tumors. Clinical Cancer Research, 13(9):2804, May 2007. 40 Chapter 5. Bibliography [9] J O’Connor, A Jackson, G Parker, and G Jayson. Dce-mri biomarkers in the clinical evalua- tion of antiangiogenic and vascular disrupting agents. British Journal of Cancer, 96(2):189, Jan 2007. [10] Cedric Bazelaire, Neil M Rofsky, Guillaume Duhamel, Jingbo Zhang, M. Dror Michaelson, Daniel George, and David C Alsop. Combined t2* and t1 measurements for improved perfu- sion and permeability studies in high field using dynamic contrast enhancement. European Radiology, 16(9):2083–2091, Sep 2006. Equation is mor. [11] L Huxham, A Kyle, J Baker, K Mcnicol, and A Minchinton. Exploring vascular dysfunction caused by tirapazamine. Microvascular Research, 75(2):247–255, Mar 2008. [12] M Penet, K Glunde, M Jacobs, and A Pathak. Molecular and functional mri of the tumor microenvironment. Journal of Nuclear Medicine, Jan 2008. [13] H Daldrup, D M Shames, M Wendland, Y Okuhata, T M Link, W Rosenau, Y Lu, and R C Brasch. Correlation of dynamic contrast-enhanced magnetic resonance imaging with histologic tumor grade: comparison of macromolecular and small-molecular contrast media. Pediatric Radiology, 28:67–78, 1998. [14] Trotter MJ, Chaplin DJ, and Olive PL. Use of a carbocyanine dye as a marker of functional vasculature in murine tumours. British Journal of Cancer, 59(5):706–9, 1989. [15] S Gwyther. New imaging techniques in cancer management. Annals of Oncology, Jan 2005. [16] R Weissleder, H C Cheng, E Marecos, K Kwong, and A Bogdanov. Non-invasive in vivo mapping of tumour vascular and interstitial volume fractions. European Journal of Cancer, 34:1448–54, 1998. [17] Tristan Barrett, Hisataka Kobayashi, Martin Brechbiel, and Peter L Choyke. Macromolecular mri contrast agents for imaging tumor angiogenesis. European Journal of Radiology, 60:353– 66, 2006. [18] K Turetschek, A Preda, V Novikov, and R Brasch. Tumor microvascular changes in an- tiangiogenic treatment: Assessment by magnetic resonance contrast . . . . JOURNAL OF MAGNETIC RESONANCE IMAGING, Jan 2004. [19] D McDonald and P Baluk. Significance of blood vessel leakiness in cancer. Cancer Research, Jan 2002. [20] Peter L Choyke. Contrast agents for imaging tumor angiogenesis: is bigger better? Radiology, 235:1–2, 2005. 41 [21] Heike E Daldrup-Link and Robert C Brasch. Macromolecular contrast agents for mr mam- mography: current status. European Radiology, 13:354–65, 2003. [22] K Li, X Zhu, D Checkley, J Tessier, V Hillier, J Waterton, and A Jackson. Simultaneous mapping of blood volume and endothelial permeability surface area product in gliomas using iterative analysis of first-pass dynamic contrast enhanced mri data. The British Journal of Radiology, 76(901):39, Jan 2003. [23] X.P Zhu. Quantification of endothelial permeability, leakage space, and blood volume in brain tumors using combined t1 and t2* contrast-enhanced dynamic mr imagine. JOURNAL OF MAGNETIC RESONANCE IMAGING, 11:575–585, 2000. [24] Catherine Westbrook and Carolyn Kaut. Mri in practice. Blackwell Publishing, 1993. [25] David L Buckley and Geoffrey J M Parker. Dcemri-oncology-chptr5. page 11, Oct 2008. [26] Geoffrey J M Parker and David L Buckley. Dcemri-oncology-chptr6. page 12, Oct 2008. [27] L Huxham, A Kyle, J Baker, K Mcnicol, and A Minchinton. Tirapazamine causes vascular dysfunction in hct-116 tumour xenografts. Radiotherapy and Oncology, 78(2):138–145, Feb 2006. [28] Kathleen Donahue and Deborah Burstein. Studies of gd-dtpa relaxivity and proton exchange rates in tissue. Magnetic Resonance in Medicine, 2005(1):66–76. [29] A Rauscher, Andreas Deistung, Jan Sedlacik, Jörg Stadler, Stephan Witoszynskyj, and Jürgen R Reichenbach. Susceptibility weighted imaging at ultra high magnetic field strengths: Theoretical considerations and experimental results. Magnetic Resonance in Medicine, 60(5):1155–1168. [30] Fernando Calamante. Dcemri-oncology-chptr4. page 15, Nov 2008. 42 Appendix A Appendix A.1 Susceptibility Equation Derivation Here we derive the susceptibility equation ( [2]) used to experimentally determine the suscepti- bility of contrast agent (Galbumin in our case), ∆B = ∆χ 2 ( cos2(θ)− 1 3 ) B0 (A.1) i.e. for θ=0, ∆B = χB0 3 (A.2) ω = γ∆B (A.3) ∆φ = −γ∆BTE (A.4) ∆φ = γ χB0 3 TE (A.5) (A.6) and finally, for θ &= 0 ∆B = ( cos2(θ)− 1 3 ) ∆χB0 TE (A.7) ∆χ = − 2∆φ B0TEγproton 1( cos2(θ)− 13 ) (A.8) 43 A.2. IDL Program to Calculate DSC Parameter Maps A.2 IDL Program to Calculate DSC Parameter Maps ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ; Creating DSC Image Maps ; ; DSC_MAP_Firas.pro ; ; ; ; April 1st, 2009 ; ; Programmed by: Firas Moosvi ; ; Assisted by: Stefan Reinsberg ; ; Applied on Data set: 30Dec08 + 31Dec08 ; ; Feridex Susceptibility ; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;/// Turn the program into a routine pro dsc_map_firas,scanid=scanid ;/// Load scan data into a matrix ’a’ ;/// Create 2 float arrays called map and map2 of size 64 ;/// These arrays hold the final map data a=mm_load_scan(scanid=scanid) map=fltarr(64,64,4) map2=fltarr(64,64,4) ;/// Begin 3 for loops spanning x,y,z ;/// z is the number of slices, in this case 2 ;/// Position of z as the last loop is important for x=0,63 do begin for y=0,63 do begin for z=0,1 do begin ;/// take three averages of each pixel, given by (x,y) for each z ;/// the first average is prior to the injection (beginning) ;/// the second average is some time after the injection (middle) ;/// the third average is the intensity average at the end 44 A.2. IDL Program to Calculate DSC Parameter Maps avg1 = mean(a[x,y,z,1:5]) avg2 = mean(a[x,y,z,15:25]) avg3 = mean(a[x,y,z,35:*]) ;/// This portion of the program allows you to have an image with 4 sections ;/// Section 1 and 2 are for z=1 and z=2 for avg2/avg1 ;/// Section 3 and 4 are the same as above except avg2/avg3 ;/// 1 - Avg2/Avg1 = 1 - Middle / Beginning -----> Large Dip = More white (closer to 1) ;/// 1 - Avg3/Avg1 = 1 - End / Beginning / -----> Large Recovery = More white (closer to 1) ;/// map[x,y,z]=((1-avg2/avg1) > 0)<1 ;/// map[x,y,z+2]=((1-avg3/avg1) > 0)<1 ;/// Here we try a map test to see if the Feridex is leaking out ;/// 1 - (Middle / End) -------> No difference between middle and end = more white (closer to 1) map2[x,y,z] = ( (1-avg2/avg3) >0)<1 ;/// print, map[x,y,z] print, map2[x,y,z] endfor endfor endfor ;/// create colour bar scale=rebin(transpose(findgen(32)/32),5,32) help,scale map2[5:9,16:47,0]=scale ;/// set window size window,0,xs=512,ys=512 ;plot 4 images, two from each z for z=0,1 do begin tvscl, rebin(map2[*,*,z],256,256),z endfor for z=2,3 do begin tvscl, rebin(map2[*,*,z],256,256),z end 45 A.3. Sample Staining Worksheet A .3 S a m p le S ta in in g W o rk sh ee t D a te : 1 2 -A p r- 0 9 P u rp o s e : C D 3 1  &  C o ll  I V  s ta in in g F o r u s e  w it h  c a p ill a ry  g a p  s ta in in g  s y s te m 5 p a ir s  o f s lid e s R o o m  t e m p e ra tu re  w it h  s h a k e r (4 0 0  r p m , 1 .5  c m ) 1 6 0 µ l/ p a ir c u t 1 0  µ m  c ry o s e c ti o n s + P ri m a ri e s d ry  o v e rn ig h t + 7 7 8 .4 µ l P B S + 0 .1 %  t w e e n 1 : 2 0 0 4 µ l ra t a n ti  C D 3 1  ( 1 :1 0 ) 5 0 % M e O H /5 0 % a c e to n e 1 0 1 : 5 0 0 1 .6 µ l ra b b it  a n ti  C o ll IV R in s e  i n  P B S  3 x 1 0  s e c + 1 : 5 0 1 6 µ l G o a t S e ru m W a s h  ( P B S  w it h  0 .1 %  T w e e n  2 0 ) w it h  s ti rr in g 2 0 S e c o n d a ri e s P u t in  c a p ill a ry  s y s te m 7 7 8 .4 µ l P B S + 0 .1 %  t w e e n 1 : 2 0 0 4 µ l a n ti  r a t A le x a  6 4 7 p ri m a ri e s  6 0 m in 1 : 5 0 0 1 .6 µ l a n ti  r a b b it  A le x a  5 4 6 R in s e  ( P B S  w it h  0 .1 %  T w e e n  2 0 ) 3 x 1 0  s e c + 1 : 5 0 1 6 µ l G o a t S e ru m s e c o n d a ri e s  6 0 m in R in s e  ( P B S  w it h  0 .1 %  T w e e n  2 0 ) 3 x 1 0  s e c + Im a g e + N o te s : 46 A.4. Feridex Staining Report A.4 Feridex Staining Report SURGICAL PATHOLOGY - HISTOLOGY Date: STAINING MANUAL - MINERALS AND PIGMENTS Page: 1 of 2 I RON -  PRUS S I A N BLUE RE A C T ION -  MA LLORY'S  ME THOD PURPOSE: To demonstrate ferric iron in tissue sections. Small amounts of iron are found normally in spleen and bone marrow. Excessive amounts are present in hemochromatosis, with deposits found in the liver and pancreas, hemosiderosis, with deposits in the liver, spleen, and lymph nodes. PR INC I P LE: The reaction occurs with the treatment of sections in acid solutions of ferrocyanides. Any ferric ion (+3) in the tissue combines with the ferrocyanide and results in the formation of a bright blue pigment called 'Prussian blue" or ferric ferrocyanide. CONT ROL: A known positive control tissue. F IXA T I V E: 10% formalin T ECHN IQUE: Cut paraffin sections 4µ. EQU I PMENT: Microwave oven, acid-cleaned glassware, non-metalic forceps. RE A GENT S: 5% P o t a s s i u m  F e r r o c y a n i d e: Potassium ferrocyanide 25.0 gm Distilled water 500.0 ml Mix well, pour into an acid-cleaned brown bottle. Stable for 6 months. C AUT ION: Lo w  t ox i c i t y  i f  no t  he a t ed.. Nu c l e a r - f a s t  Re d: See Retic 5% Hy d r o c h l o r i c  A c i d: Hydrochloric acid, conc. 25.0 ml Distilled water 475.0 ml Mix well, pour into brown bottle, stable for 6 months. C A UT ION: Co r r o s i v e ,  a v o i d  c on t a c t  a nd i n h a l a t i o n . Wo r k i ng  S o l u t i on: 5% potassium ferrocyanide 25.0 ml 5% hydrochloric acid 25.0 ml Make fresh, discard after use. C AUT ION: A vo i d  con t a c t  and i nha l a t i on. 48 A.4. Feridex Staining Report MINERALS AND PIGMENTS IRON Page: 2 of 2 S A FETY: Wear gloves, goggles and lab coat. Avoid contact and inhalation. Potassium ferrocyanide; Low toxicity as long as it is not heated, it will release cyanide gas. Hydrochloric acid; target organ effects on reproductive system and fetal tissue. Irritant to skin eyes and respiratory sytem. PROCEDURE: 1. Deparaffinize and hydrate to distilled water. 2. *Working solution, * microwave, 30 seconds. Allow slides to stand in solution for 5 minutes, in the fume hood. 3. Rinse in distilled water. 4. Nuclear-fast red, 5 minutes. 5. Wash in tap water. 6. Dehydrate, clear, and coverslip. *Conventional method: room temperature for 30 minutes. RESULTS: Iron (hemosiderin) blue Nuclei red Background pink REFERENCES: Sheehan D, Hrapchak B, Theory and practice of Histotechnology, 2nd Ed, 1980, pp217-218, Battelle Press, Ohio Luna L, Manual of Histologic Staining Methods of the AFIP, 3rd Ed, 1968, pp 183, McGraw-Hill, NY Crookham,J, Dapson,R, Hazardous Chemicals in the Histopathology Laboratory, 2nd ED, 1991, Anatech Prepared:                                        By: Approved:                                       By: Downloaded from WebPath: Internet Pathology Laboratory http://www-medlib.med.utah.edu/WebPath/webpath.html 49 Physics 449 Report: Feridex Mapping April 14, 2009 1 Mapping Feridex histologically Feridex (generically ferrumoxide), is a sterile aqueous colloid of superparamagnetic iron oxide (SPIO) asso- ciated with dextran. Administered intravenously, it is an MR contrast agent usually used for the detection of liver lesions. Feridex is taken up by macrophages, found only in healthy liver cells but not in most tu- mors. We want to use Feridex in the DSC weighted portion of our protocol, alongside Gd-DTPA as the DCE portion. The problem with using Feridex so far (instead of Galbumin) is tracking and/or mapping the movement of Feridex as it extravasates out of tumour vessels. 1.1 Magneto-capsules - Dextran-specific FITC-conjugated antibody • LIM-Sun Method, described below in detail from the Barnett paper. Magnetoencapsulation. Magnetocapsule synthesis is based on a one-step modication (that is, Feridex addition) of the Lim-Sun method29. Our modication uses an electrostatic (van de Graaff ) droplet generator, which produces smaller, stronger and more uniform capsules as compared with 2 50 Physics 449 Report: Feridex Mapping April 14, 2009 Table 1: Description of Feridex Description Value r1,r2,Bo 40.0,160,0.47T Concentration 11.2 mg Fe/mL Avg. Formula FeO1.44 Active Ing. 11.2 mg of Fe Inactive Ing. 1 61.3 mg of mannitol Inactive Ing. 2 5.6-9.1 mg/mL dextran Inactive Ing. 3 0.25-0.53 mg/mL citrate Hydrodynamic Diameter 120-180 nm *(80-150 nm also cited) Size of Crystal Core 5.55 nm Coating Dextran T10 kDa the older air-jet technique. Before encapsulation, human cadaveric islets were passed through a 20-g needle. We suspended cells, adjusted to 400 islet equivalents per ml or 1.5 E7cells/ml (bTC-6), in 2% (wt/vol) ultrapuried sodium Protanal-HF alginate (FMC Biopolymers) and 20% (vol/vol) Feridex (Berlex Laboratories). We passed this solution through a needle at 200 ml/min using a nanoinjector pump. We collected droplets, representing islet cells surrounded by the rst layer of alginate, in a Petri dish containing 100 mM CaCl2 in 10 mM HEPES, and washed them three times. We suspended gelled droplets in 0.05% poly-L-lysine (2224 kDa; Sigma) for 5 min to crosslink alginate and Feridex. We washed and resuspended droplets in 0.15% Keltone HVCR alginate (Monsanto) for 5 min, and then washed them again. For capsule rupture, we manually agitated magnetocapsules in a 50-ml conical tube lled with 1-mm glass beads. 1.2 Alexa-647 conjugated Feridex • Refer to as the paper for the full methods of creating this • Preparing Alexa Fluor 647 conjugated Feridex • Feridex added to a solution containing KOH, (ddH2O), epichlorohydrin • Reacted for 12 hours with constant shaking • Concentrated ammonia added to the Feridex and reacted overnight at 37C • Feridex was reacted with 1 mg of Alexa Fluor 647 succinimidyl ester overnight at room temperature • Centrifuged and supernatant removed • Assuming an average particle diameter of 80 nm, we determined that there were approxi- mately 140 molecules per iron particle More descriptive protocol Preparation of Fe[647] and Fe[750] Nanoparticles Synthesis of Alexa Fluor 647- and Alexa Fluor 750-conjugated Feridex (Fe[647] and Fe[750], respectively) was based on 3 51 Physics 449 Report: Feridex Mapping April 14, 2009 previ- ously published methods [15]. Briefly, 1 ml of Feridex (11.2 mg/ml) was added to a solution containing 1.6 ml of KOH, 0.7 ml of double-distilled H2O (ddH2O), and 0.7 ml of epichlorohydrin [19]. The mixture was reacted for 12 hours with constant shaking. To produce reactive amines on the dextran coat, concentrated ammonia (0.5 ml) was added to the Feridex and reacted overnight at 37C. Excess epichlorohydrin and ammonia were removed by extensive dialysis against ddH2O using 12,000 14,000 molecular weight cut off tubing. Feridex was reacted with 1 mg of Alexa Fluor 647 or 750 succinimidyl ester (Molecular Probes, Eugene, OR, http://probes. invitrogen.com) overnight at room temperature. Excess fluorophore was removed by centrifuging the sample at 160,000g for 30 min- utes. The supernatant was discarded, and the pellet was resuspended in phosphate-buffered saline (PBS) buffer (pH 7.4). The centrifu- gation step was repeated four times to ensure that unconjugated fluorophore was removed from the sample. Removal of excess fluoro- phore was confirmed using a fluorometer. To disperse and remove large nanoparticle aggregates, the sample was sonicated for 5 minutes and filtered using a 0.2 M size-exclusion filter. The iron concentration was measured using the method de- scribed by Stookey [20]. The number of fluorescent molecules was calculated by using a standard curve of known Alexa Fluor 750 concen- tration. From these measurements, assuming an average par- ticle diameter of 80 nm, we determined that there were approximately 140 molecules per iron particle. 1.3 Prussian Blue - Fe staining • To demonstrate ferric iron in tissue sections • The reaction occurs with the treatment of sections in acid solutions of ferrocyanides • Tissue stained and positive control • Paraffin sectioning 5µm thick • Ferric iron is blue and nuclei are red • See attached protocol for Prussian Blue stain. 1.4 Radio-labeled Feridex • Wasn’t able to find situations of radiolabelled Feridex • I have a couple of papers that describe the structure of Feridex. • Could also use this to directly attach FITC tags to Feridex • I didn’t really understand them, but it might to someone that knows a bit more chemistry • Still have to read AK’s paper for the radio labelling protocol. 4 52 Glossary blood brain barrier The blood brain barrier is a special feature (many tight junctions) of brain tissue as it restricts the passage of molecules of a certain size between the blood stream and brain tissue.. 4, 15 carbocyanine Carbocyanine (DiOC7(3)) is a small molecular dye commonly used in histology to measure perfusion. Because of its small size, carbocyanine stains cells immediately adjacent to blood vessels outlines tumour vasculature. The dye commercially available (Invitrogen, SKU# D-378) fluoresces at 488nm and can be imaged using the digital microscope at the BC Cancer Center.. 5, 6 CD31 An endothelial cell marker that when stained for, marks blood vessels.. 5, 20, 23 Contrast Agent Radioactive, fluorescent or magnetic chemical compounds that aid in the vis- ibility of internal bodily structures in imaging. In this study, we will be using Galbumin (Gadolinium atoms attached to the protein Albumin) as the contrast agent.. 3 dextran Complex sugar. In context, dextran is used to increase the effective size of the target molecule (Gd-DTPA in this case. 35, 37 Diffusion This is the traditional, classical method of molecules moving according to brownian motion.. 4, 5 EES Extravascular extracellular space - referred to the space that isn’t occupied by vasculature (vessels) or cells. 16 endothelium A layer of cells that line the interior of blood vessels.. 4 extravasate To leak out of... In context, extravasation is a term used to denote the leakage of fluid from a container.. 17, 36 extravasation In general, this term refers to the leakage of a fluid out of its container. In the context of this study, it’s the leakage of molecules (contrast agents, drugs, markers etc.) out of the blood vessel and into the extra cellular matrix.. 4, 5, 37 hoechst A small fluorescent biomarker that is commonly used to mark perfusion (from blood vessels). 5, 23 53 Glossary hydrostatic Hydrostatic - fluids at rest. In context, hydrostatic pressure is the force per unit area delivered from water that is static (in equilibrium and not moving). 5 immunohistochemical Immunohistochemistry is the process of localizing biological molecules (often proteins) at the cellular level using histology (study of thin sections, usually frozen) immunology (antibodies binding to antigens) and chemical principles.. 3, 6 immunohistochemistry Immunohistochemistry is the process of localizing biological molecules (often proteins) at the cellular level using histology (study of thin sections, usually frozen) immunology (antibodies binding to antigens) and chemical principles.. 3, 4 in-vivo Tumours will be xenografted in NOD SCID mice of the cell type HCT116 and HT29. 8 interstitial Area that surrounds cells and vessels filled with fluid.. 5, 16, 34 intravascular Stays inside blood vessels. In contrast, large molecular weight contrast agents stay confined to within blood vessels and do not extravasate.. 37 metastasis Metastasis is referred to as the spread of disease (usually cancer) from one part of the body to another non-bordering part. One method of metastasis is cancer cells moving through the bloodstream from one part of the body to another.. 4 Mouse Type NODSCID - Non-obese diabetic, severe combined immunodeficiency mice. 5 necrosis Group (or area) of cells that is dying or dead. Some tumours show patches of necrosis in areas distant from blood vessels (nutrients) and these areas are essentially empty space.. 29, 34 tight junctions Tight junctions are the closely associated areas of two cells whose membranes form an impermeable barrier to fluid. In context, generally the number of tight junctions in tissue is related to the permeability of the tissue.. 4 vascular networks A vascular network is the arrangement of blood vessels of a particular organ or tissue.. 4 vasculature The network of blood vessels of an organ or body part and includes distribution of all vessels, including arteries, capillaries and veins.. 3, 5, 20 54

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 33 0
Japan 6 0
China 6 2
Australia 2 0
Canada 1 0
United Kingdom 1 0
City Views Downloads
Washington 24 0
Tokyo 6 0
Beijing 4 0
Mountain View 4 0
Shenzhen 2 2
Baltimore 2 0
Sydney 2 0
Las Vegas 2 0
Unknown 1 9
Vancouver 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items