Quantitative multiparametric imaging for the evaluation of nasopharyngeal carcinoma using PET and DCE-MRI

Nasopharyngeal carcinoma (NPC) is an aggressive head and neck cancer ranked as the 5th most common in Hong Kong. We aimed to study the role of dynamic contrast-enhanced MRI (DCE-MRI) and dynamic 2-deoxy-2-[fluorine-18]fluoro -D-glucose positron emission tomography (FDG-PET) for characterizing NPC tumors in newly-diagnosed patients, and to quantitatively evaluate the intratumoral heterogeneity of NPC. In Chapter 2 we employed semi-quantitative analysis of DCE-MRI to study the dynamic enhancement pattern by analyzing the time-intensity curves in 25 NPC patients. Our findings suggested that high blood flow caused a high initial intensity enhancement rate (ER), and that neovasculature due to tumor angiogenesis in tumors of larger volume or higher T-stage caused more accumulation of contrast agent which can be detected by DCE-MRI. PET and semi-quantitative DCE-MRI parameters were not correlated and may reflect different physiological/molecular processes in the microenvironment of NPC tumor. However the major limitation of semi-quantitative analysis was that the physiological correlates of these parameters were unclear. In Chapter 3 we applied quantitative analysis of DCE-MRI to study the permeability and perfusion characteristics in the same cohort as in Chapter 2. Our findings implied that the permeability may be high compared to blood flow in NPC tumor. We also observed significant correlations between iAUC (the initial area under the time-intensity curve) by semi-quantitative analysis and ve (the volume fraction of extravascular extracellular space) by quantitative analysis, and between the two rate constants (kep’s) from these two methods, which showed that semi-quantitative analysis was a feasible alternative in reflecting the physiological characteristics of NPC. However, we did not observe any significant correlation between PET and DCE-MRI quantitative parameters, also suggesting that PET and DCE-MRI reflected different physiological information in NPC. In Chapter 4 we applied dynamic PET scan to study the glucose metabolism in 18 NPC tumors (16 included in DCE-MRI cohort). Our findings showed that the overall FDG uptake was mainly composed of the FDG in tissue compartment (Ki), which was governed by the phosphorylation (k3) but not the transport of FDG (K1). This finding may further indicate a potential role of the phosphorylation rate k3 in NPC. Dynamic PET parameters did not correlate with DCE-MRI, indicating that the two modalities reflect different molecular information in NPC. In Chapter 5, intratumoral heterogeneity in NPC tumors of 40 patients was studied using 18F-FDG PET scan. Our findings showed that as tumors grew to a larger volume and higher T-stage, they showed more heterogeneous glucose metabolism. It was found that more heterogeneous tumor was associated with worse disease-free survival, indicating that tumor metabolic heterogeneity may play an important role for NPC patient prognosis. To summarize, these results showed that DCE-MRI and dynamic PET improved our understanding about the physiological/molecular process of NPC, and, these two modalities reflected different physiological information in the microenvironment of NPC tumors. This indicated that the relationship between supply of nutrients such as glucose and blood flow/permeability is complex and not directly related. Moreover, intratumoral heterogeneity by PET scan was also of importance in prognostication.published_or_final_versionDiagnostic RadiologyDoctoralDoctor of Philosoph