Doctor of Medicine and Doctor of Philosophy (MDPhD)
Targeting the Tumour Microenvironment for PET Imaging and Radionuclide Therapy
The full abstract for this thesis is available in the body of the thesis, and will be available when the embargo expires.
The development of novel bioconjugates; antibodies, antibody mimetics and peptides is revolutionising the targeted molecular imaging approach to oncology. Many bioconjugates with high and specific uptake in tumours are transitioning into clinical oncology imaging. Monoclonal antibodies are returning to diagnostic imaging due to the clinical impact and commercial success of many disease-modifying monoclonal antibodies and antibody-drug conjugates. One hindrance to mAb based imaging is their long biological half-life; requiring days to visualise high tumour-to-non-target contrast. A variety of avenues to improve the contrast and clearance of circulating mAbs are in development. Radiometals that are well suited for imaging antibodies as well as antibody mimentics or peptides are becoming more in demand as clinical translation occurs. Current constraints and requirements for conventional radiometal production can limit the accessibility to certain cyclotron centers and may not be easily integrated into existing cyclotron centers. Efforts have been made to modify conventional radiometal production by introducing a liquid target approach. Limitations with yield and purification methods are hindering the feasibility of producing radiometals with a liquid target. Herein, we propose to produce radiometals with a liquid target and compare yields, optimize purification procedures to ensure high radiolabeling yields and assess their usefulness to radiolabel existing and novel bioconjugates. Additionally, we use a variety of strategies to improve tumour-to-non-target contrast ratios and pharmacokinetics of mAbs which can be applied to imaging of novel mAbs.We demonstrate the ability to produce and purify ⁸⁹Zr with both a solid and liquid target with sufficient yields for radiolabeling of antibodies. We demonstrate that liquid target produced ⁸⁹Zr is a suitable alternative approach to conventional solid target ⁸⁹Zr by demonstrating suitable antibody radiolabeling yields for in vivo imaging. We demonstrate that liquid target produced ⁶⁸Ga can be separated effectively from the staring zinc material and from other metal impurities. This successful purification enabled high radiolabeling of DOTATOC. The radiolabeled ⁶⁸GaDOTATOC from liquid target ⁶⁸Ga was compared with an in vivo study to that of conventional generator ⁶⁸Ga. We demonstrate the feasibility of producing, purifying and radiolabeling liquid target ⁶⁸Ga for in vivo imaging.
Many compounds mimicking endogenous molecules have been used as a starting point to develop targeted diagnostic and therapeutic radiotracers. In particular, radiolabeled peptidomimetics, in association with positron emission tomography combined with computed tomography (PET/CT), are powerful tools to detect cancer with high sensitivity. Peptide-based radiotracers have the advantage of combining favorable pharmacokinetics that allow the use of short-lived isotopes, with a flexible modular design that offers a high versatility for functionalization, making them optimal for developing targeted imaging probes. The bradykinin receptors, which are powerful mediators of inflammation, have been shown to be highly expressed in many common cancers, notably breast and prostate cancers. The purpose of this project was to evaluate the human Bradykinin Receptor 1 (hB1R) as a potential target for cancer imaging and radionuclide therapy. Analogs of [des-Arg¹⁰] Kallidin (KD) were synthesized and labeled with ⁶⁸Ga or ¹⁸F. Following determination of their affinity for hB1R, selected tracers were evaluated in vitro and in vivo using hB1R expressing cells to select optimal radiotracers to imaging by positron emission tomography. The replacement of key amino acids at peptidase cleavage points by unnatural aminoacids improved the stability of the radiolabeled [des-Arg¹⁰]KD analogs in vitro and in vivo. Such peptides were used successfully for h1BR imaging by PET/CT in preclinical models. The use of hydrophilic and in particular cationic linker significantly improved tumour accumulation of various bradykinin analogues. Tracers combining the most favorable features gave high tumour to normal tissue contrast, by combining specific and high tumour uptake with low background and rapid clearance. The accumulation of agonist and antagonist radiotracers in tumours was also compared. In summary, we developed several promising bradykinin receptor ligands, as radiolabeled probes for cancer imaging.
Tumour hypoxia has long been recognized as an impediment to radiotherapy and chemotherapy. Cancers that are hypoxic tend to be aggressive, with high propensity for distant metastasis. As hypoxia is a salient feature of most solid cancers, targeting components of the hypoxia-induced signaling cascade has been proposed as a means for oncologic treatment. The key enzyme mediating hypoxia-induced stress response in cancers is carbonic anhydrase IX (CA-IX). Regulated by hypoxia-inducible factor 1α (HIF-1α), CA-IX catalyzes the reversible hydration of carbon dioxide to bicarbonate ion. CA-IX promotes cancer cell survival by transporting bicarbonate ions into the cell to maintain pH homeostasis during glycolysis. CA-IX is well-established as a surrogate marker for cellular hypoxia. Overexpression of CA-IX has been observed in a broad spectrum of cancers including: breast, cervix, ovarian, bladder, brain, colon, lung, kidney, head and neck cancers. In healthy individuals, CA-IX is expressed at low levels except in the gastrointestinal tract where it is involved in the process of cell differentiation. As CA-IX is pathologically expressed by cancer cells and located at the cell surface, it has emerged as a promising imaging/therapeutic target.In this thesis, we communicate the development of molecular antigen recognition molecules as potential radiotracers for CA-IX targeted nuclear imaging. We identified two classes of sulfonamide derivatives that successfully delineated CA-IX expression in tumour-bearing mice. Isoform selectivity, the major challenge for small molecule inhibitor-based imaging, was achieved via a multivalent approach or by conjugating pharmacophores to polyaminocarboxylate chelators. With good tumour-to-nontarget ratios and fast pharmacokinetics, some of these agents warrant further investigation as surrogate hypoxia imaging agents. Additionally we radiolabeled three novel monoclonal antibodies (mAbs) and one affibody for CA-IX imaging, with one mAb in particular showing significant accumulation in tumours. Collectively, this research provides a non-invasive platform to characterize and quantify expression of CA-IX in primary lesions and across metastatic sites. The diagnostic information can be readily integrated with emergent pharmaceuticals to increase effectiveness and safety of CA-IX or hypoxia-directed treatments for cancer patients.
In the past years, peptide based radiopharmaceuticals have turned into favorable molecular imaging agents for specific targeting of cancer. This is mainly because many tumors happen to overexpress certain regulatory peptide receptors. For instance, the gastrin releasing peptide (GRP) receptors are overexpressed in prostate cancer–the most common malignancy among men–and somatostatin 2a (SST2a), and neuropeptide Y1 (NPY1) receptors are overexpressed in breast cancer–the most common cancer among women. There are disadvantages to most existing imaging techniques used for early detection of prostate and breast cancer. Thus, the objective of the work presented in this thesis was to develop a novel and specific diagnostic approach using radiolabeled peptides for PET imaging to localize lesions of breast and prostate cancers. Towards this end, different derivatives of GRP, SST2a, and NPY1 peptides were synthesized and their binding affinity was confirmed in vitro. The promising candidates were radiolabeled with ¹⁸F or ⁶⁸Ga–the ideal radioisotopes in PET applications. Two different ¹⁸F labeling methods (click chemistry and trifluoroborate exchange reaction) were conducted. Finally, the biological evaluation of radiopharmaceuticals was performed in vivo by using animal models of prostate and breast cancer. In the click chemistry approach, introducing PEG spacers to GRP derivatives improved the in vitro properties and the pharmacokinetics in the prostate tumor model, leading to better tumor visualization by PET imaging. The trifluoroborate exchange reaction proved to be a superior technique–by both radiochemistry and biological criteria–in GRP labeling, resulting in an excellent tumor uptake with ¹⁸F-AmBF3-MJ9. The same approach was also successful for targeting SST2a receptors in mice bearing rat pancreatic tumor cells. The data achieved with this labeling method suggest the potential application of these radiopharmaceuticals for diagnosis in cancer patients. A high tumor to background ratio was achieved in the Zr-75-1 tumor model with ⁶⁸Ga-DOTA-TATE and ⁶⁸Ga-NODAGA-LM3. Hence, this cell line is a promising breast cancer model for SST2a imaging. The NPY1R compound ¹⁸F-ALK-BVD15 was not metabolically stable and showed low receptor-mediated tumor uptake in NPY1R-positive tumors. Different strategies will need to be explored in order to modify and improve the stability of the peptide.