Dynamic magnetic resonance imaging for tumor prognosis.

by Jiang, Lan.

DYNAMIC MAGNETIC RESONANCE FOR CANCER PROGNOSIS Publication No. ____________ Lan Jiang The University of Texas Southwestern Medical Center at Dallas, 2006 Supervising Professor: Ralph P. Mason, Ph.D., C Sic., C. Chem. Breast and prostate cancers are the most common non-smoking cancers among American women and men. Radiotherapy and chemotherapy in conjunction with surgery are the most common treatment protocols in the clinic. However, a lot of experimental and clinical studies have shown that tumor hypoxia and the microcirculation play a very important role in cancer progression and therapy. There is strong evidence that hypoxic cells are one of the major reasons for failure to control tumors with conventional radiotherapy and chemotherapy. Several approaches (hyperthermia, and carbogen inhalation), which improve tumor oxygenation during radiotherapy and chemotherapy, have been used in clinical trials. There is increasing demand for tumor prognostic information in the clinical setting. So far, increasing clinical data have indicated that poorly oxygenated tumors have poor prognosis. To better understand the underlying tumor physiological mechanisms, it is very important to develop novel non-invasive approaches to accurately assess tumor microcirculation and oxygenation for further therapy planning. However, these parameters have been extremely difficult to assess in routine clinical practice and have therefore not been easily integrated in to general patient care. VI With development of MRI the non-invasive technique, BOLD (Blood Oxygenation Level Dependent) contrast MRI, has been widely used for neuroscience research to detect brain activations. Because deoxyhemoglobin (dHbO2) is paramagnetic and oxyhemoglobin (HbO2) is non-magnetic, the change of concentration of deoxyhemoglobin and oxyhemoglobin can cause a Bulk Magnetic Susceptibility (BMS) change and the T2* signal response during MR imaging. Here, I applied this technique to assess tumor physiological characteristics. In order to study the BOLD mechanism, I designed a phantom system and built it for in-vitro study. Since inhalation of oxygen could cause variation in the blood flow and oxygenation, and BOLD MRI is sensitive to both these factors, it becomes very important to explore the correlation between the BOLD response and these two factors. Considering the different vascular orientation, the angle between vessel and the static magnetic field (B0) could be further analyzed in phantoms. The phantom study showed that contribution of oxygenation was much higher than that of flow to the BOLD signal. Interestingly, a signal decrease was observed in extra-vessel region accompany increasing intra-vessel oxygenation. The DCE (Dynamic Contrast Enhanced) and BOLD MRI have been compared in the animal experiments and the clinical setting. The experimental pre-clinical results showed that tumor sublines with different vascular development showed different DCE and BOLD regional response. This correlation between DCE and BOLD in regional response indicated the potential value of BOLD technique in tumor physiological research. VII Finally, the clinical results showed there is prognostic value in DCE and BOLD MRI. There is high correlation between high BOLD response and good therapeutic outcome. BOLD and DCE MRI provide novel non-invasive prognostic tools for the clinic. It provides new insight into tumor physiological changes during chemotherapy and radiotherapy. I found correlation between BOLD response in tumors accompanying oxygen breathing and the clinical response of advanced local breast to chemotherapy. This technique could early become a routine addition to the clinical setting benefiting cancer patient population in near future. VIII