Magnetic Resonance Imaging of Oxygen Nano-Carriers for the Treatment
of Hypoxic Tumours

Emma Bluemke

Supervised by Prof Eleanor Stride and Prof Daniel Bulte

Institute of Biomedical Engineering
University of Oxford

Informal Thesis Summary

Tumours often have very low levels of oxygen. Although at first, this might seem like a good thing — suffocate & kill the tumour, right? — it's not. It actually makes them more difficult to kill, because a lot of our treatments (like radiation therapy) require oxygen in the tissue in order to work properly. To help make cancer treatment as effective as possible, a lot of people are working on ways to increase the amount of oxygen in the tumour during treatment. My co-supervisor Professor Eleanor Stride is developing tiny, safe bubbles to help deliver oxygen to the tumours (you can watch her explain the bubbles on youtube here).

Researchers and clinicians need a method to verify whether or not the oxygen is reaching the tumour, and MRI might be a good method to use, since it's safe, non-invasive, and already used in hospitals. This is why I'm co-supervised by Professor Daniel Bulte, whose research focuses on MRI and oxygen. My thesis research focused on determining whether or not MRI is a good method to use to measure the oxygen changes from these nanobubbles. In order to answer this question, I have performed new MRI experiments, created mathematical models, and analysed medical image data from two clinical trials.

In the process of answering this thesis question focusing on nanobubble delivery, my thesis resulted in three separate mathematical models that are useful for areas of research beyond the focus of this thesis, especially low-field MRI, quantitative MRI, and oxygen-enhanced MRI. I have hosted the open-source code for each model publicly on GitHub for future researchers to use, adapt, and improve upon easily.

Chapter Overview

Chapter 1: Thesis Introduction

This thesis aims to examine whether clinically available MRI techniques can be used to detect oxygen delivery from nanobubbles developed for the treatment of tumour hypoxia. In this chapter, the motivation of this thesis research is introduced by providing background information regarding the problem of tumour hypoxia and the development of oxygen carriers to aid tumour therapy. The challenge of measuring the oxygen delivery from these carriers is discussed, and a review of possible imaging-based methods that could be used to measure the oxygen delivery is provided.

This chapter includes edited material from the following forthcoming manuscript: 

  • E Bluemke, E Stride, D Bulte. A Review of Imaging Methods for Measuring Oxygen Delivery. In preparation.

Chapter 2: Oxygen Relaxivity in Nanobubbles 

To address the question of whether the change in oxygen caused by the nano-carriers can be detected using clinically available MRI techniques, a first step was to establish the limit of detection of oxygen via the available MRI methods. Towards this aim, we performed two phantom MRI experiments measuring the relaxivity and limit of detection in oxygen nanobubble solutions in both 3 Tesla and 7 Tesla MRI. 

This chapter includes edited material from the following publication: 

Subsets of these results were presented at the following conferences:

Chapter 3: Pre-Clinical Pilot Experiments

Chapter 2 presented the relaxivity and limit of detection of oxygen via measurement of R1 (a relaxation measurement from MRI), suggesting that it could be a feasible technique to detect oxygen delivery. This chapter presents the results of three pilot preclinical experiments to test whether the oxygen delivery from the nanobubble injection was detectable using R1 measurements. 

Although the results of these experiments were inconclusive, they inspired investigations that exposed important possible confounding factors such as temperature, and most importantly, these contradictory results were the motivation for the quantitative model presented in the subsequent chapters.

This chapter includes material that was presented at the following conference: 

Chapter 4: A Model of Oxygen Relaxivity in Bodily Fluids

In order to model the expected R1 response from nanobubble oxygen delivery, a value for the relaxivity in bodily fluids must be known. In this chapter, all reported relaxivity values from the literature in phantoms, saline and water solutions, and vitreous fluid are consolidated, which showed a large variability in values of relaxivity. Therefore, a simplified model is proposed for estimating relaxivity in these fluids based with respect to the magnetic field and temperature, which is able to explain the majority of the variance in relaxivity values.

This chapter includes edited material from the following publication: 

Chapter 5: A Model of Blood R1 in Response to Hyperoxia

In order to model the expected R1 response from nanobubble oxygen delivery, the expected R1 change in blood following an increase in oxygen must be known. Therefore, in this chapter, a general model to calculate the R1 of blood is presented, accounting for haematocrit, oxygen saturation, oxygen partial pressure, and magnetic field strength under hyperoxic conditions.

This chapter includes edited material from the following publication: 

A subset of these results was presented at the following conference:

Chapter 6: A Model of Tissue R1 Response to Oxygen Delivery 

Although the model in Chapter 5 accurately describes the expected R1 change in blood, the aim of this thesis is to detect the oxygen delivery from these nanobubbles in tissue. Therefore, in this chapter, a 3-compartment model is proposed for estimating the changes in R1 that could be expected in tissues depending on field strength, blood oxygen saturation, blood volume, hematocrit, oxygen extraction fraction, and changes in the partial pressure of oxygen.

This chapter includes edited material from the following publication and forthcoming manuscript: 

A subset of these results was presented at the following conference:

Chapter 7: Conclusions, Clinical Translation, and Future Work

In this chapter, the major findings from each chapter of this thesis are summarised. These estimates are, however, based on theoretical calculations, and ultimately, successful clinical translation is the end goal for the therapeutic nanobubbles discussed in this thesis. Although it was not possible to perform clinical experiments with nanobubble delivery within the scope of this thesis, the most relevant adjacent research area is clinical OE-MRI, and it was possible to analyze the results from two recently completed clinical OE-MRI studies. These results are useful for shedding light on several elements that would be relevant to using this method in a clinical experiment involving nanobubbles, such as the benefits and drawbacks of a chosen T1 mapping method, vulnerability to patient motion, and potential image artefacts. In this chapter, these original research results are used to draw conclusions about the feasibility of using R1 measurements to monitor oxygen delivery from nanobubbles in a clinical MRI setting. Lastly, the broader potential applications of the work produced for this thesis are discussed.

This chapter includes edited material from the following publications and forthcoming manuscripts:

A subset of these results was presented at the following conference: