Improving Cancer Biomarkers through Magnetic Resonance Technology
by
Matthew Luke Zierhut
Cancer is a complex disease that is not easily characterized, especially in its early stages. However, cancer does demonstrate numerous abnormal traits compared to healthy tissue of similar origin. These traits are most commonly observed in a tissue sample obtained through a surgical biopsy. Biopsies are not available for all tumors, and when performed, they can be highly invasive procedures.
Nuclear magnetic resonance (NMR) allows for techniques that can non-invasively investigate several abnormal characteristics common to cancer. MR imaging (MRI) can be used to see contrast in cancerous tissue with different MR properties and to monitor kinetic changes associated with cancer progression from an exogenously injected compound. MR spectroscopic imaging (MRSI) can also be used to assess the amount of certain endogenous substances, some of which are noticeably altered in the presence of cancer. Furthermore, dynamic NMR spectroscopy and MRSI can enable metabolically active exogenously injected compounds to be monitored, along with their metabolic products; thus, both kinetic and metabolic properties of cancer can simultaneous be ascertained.
In this thesis project, three techniques based on NMR technology were investigated to determine their value in assessing cancer characteristics. Kinetic modeling of T1 weighted and T2* weighted dynamic contrast enhanced 1H MRI (DCE-MRI) were compared to evaluate which technique was more clinically valuable. Various techniques for reducing acquisition time for three-dimensional 1H MRSI were also compared to determine which technique produced clinically relevant data with the fewest artifacts in the human brain. Finally, kinetic modeling was applied to dynamic 13C spectroscopy data in rats, mice, and dogs to evaluate its promise as a future clinical technique.
The results from this research project show that kinetic modeling of T1 weighted DCE-MRI data and 1H 3D echo-planar spectroscopic imaging (EPSI) techniques are particularly valuable for cancer research, especially when compared to methods aiming for similar results. Additionally, by applying kinetic modeling and EPSI techniques to hyperpolarized 13C data, it is possible to non-invasively monitor cancer metabolism in vivo. Therefore, implementing hyperpolarized 13C1-pyruvate techniques in humans may drastically improve cancer prognoses in the near future.
