Search

In this dissertation, the qualitative and quantitative multimodal imaging possibilities of holmium-166 loaded poly(L-lactic acid) microspheres (166Ho-PLLA MS) are explored and exploited to improve biodistribution assessment and dose calculations for planning, image-guidance and evaluation of hepatic arterial radioembolization of liver malignancies. This relatively novel radiotherapy treatment modality represents a promising treatment option for patients with unresectable liver malignancies. In hepatic arterial radioembolization, radioactive microspheres are administered directly into the hepatic artery using a catheter. Targeting of tumors is accomplished by exploiting the predominance of the arterial blood supply to liver tumors, while normal parenchyma largely depends on portal blood supply. Ideally, this results in a high tumor-to-liver ratio, leading to an increased radiation dose to the tumor tissue while minimizing exposure to healthy liver parenchyma. An essential element of successful radioembolization of hepatic malignancies is preprocedural biodistribution assessment. The large variation in vascularity of tumor and liver tissue observed between patients necessitates extensive treatment planning to assure a favorable dose distribution in each individual patient. Due to their multimodal imaging properties, 166Ho-PLLA MS are believed to be an improvement as compared to the already clinically applied yttrium-90 microspheres, which lack high quality medical imaging possibilities. To fully exploit 166Ho-PLLA MS’ imaging opportunities and enable accurate biodistribution assessment and dose calculations, the qualitative and quantitative multimodal imaging possibilities of 166Ho-PLLA MS were investigated, focusing on magnetic resonance imaging (MRI), X ray computed tomography (CT) and Single Photon Emission Computed Tomography (SPECT). In vitro experiments demonstrated that SPECT has the highest sensitivity and lowest detection limit for 166Ho-PLLA MS, followed by MRI and CT, respectively. The development of highly-loaded holmium microspheres, with increased holmium content from 17% to 45% w/w, strongly increased the specific activity for therapeutic purposes and more than doubled the multimodal diagnostic properties for SPECT, MRI and CT. MicroCT imaging was demonstrated to enable high resolution 3D biodistribution assessment of 166Ho-PLLA MS in liver tissue after hepatic arterial radioembolization of a Vx2 tumor-bearing rabbit liver both qualitatively and quantitatively. Single microspheres were detectable using microCT. Microspheres mainly lodged in the periphery of the tumor, revealing a highly skewed cluster volume distribution towards small volumes. Macroscopically, MRI was demonstrated to enable selective depiction of Ho-PLLA MS with positive contrast in liver tissue in the presence of macroscopic magnetic field distortions, using a method called susceptibility gradient mapping. For MR-based quantitative assessment of the distribution of Ho-PLLA MS in liver tissue using T2* relaxometry, sampling of free induction decay (FID) was shown to be superior to sampling of the spin echo (SE), due to diffusion sensitivity of the SE signal decay time course. By means of MR experiments and Monte Carlo simulations FID signal behavior of a diffusive medium containing Ho-PLLA MS was shown to exhibit monoexponential signal decay. Both the static dephasing theory and the theory of strong field behavior accurately predicted transverse relaxivity, allowing MR-based quantification of the local concentration of HoMS. Furthermore, a method which increases the upper detection limit of the Ho-PLLA MS concentration by estimating the S0 value of the signal decay curve to be used in the quantitative fitting procedure was proposed. Finally, the feasibility to utilize quantitative MR data for dosimetric calculations of 166Ho-PLLA MS was demonstrated in an anthropomorphic gel phantom, indicating the potential of MR-based dosimetry for planning, guidance and evaluation of transcatheter radioembolization of hepatic malignancies.

In this thesis, the value of various MRI techniques was addressed in patients with nonpalpable breast lesions. Both studies with an emphasis on the clinical – epidemiological aspects of MRI of the breast (chapters 2, 3, 6) and studies that focused primarily on the physics – technical aspects (chapters 4, 5, 7) were performed. In a diagnostic meta-analysis (chapter 2), we found an overall sensitivity of dynamic contrast-enhanced Breast MRI of 0.90 (95% CI 0.88 – 0.92) and an overall specificity of 0.72 (95% CI 0.67 – 0.77) using the bivariate analysis apporach for diagnostic meta-analyses. The specificity varied more across the individual studies than the sensitivity did. In chapter 3, the rationale, objectives and study design of the MONET - study (MR mammography Of Nonpalpable BrEast Tumors) were described. The purpose of this randomized study is to evaluate whether performing dynamic contrast-enhanced Breast MRI in addition to mammography in patients with nonpalpable suspicious breast lesions will improve breast cancer management. The results of the MONET – study are expected to be available medio 2010. The influence of the choice of b-values on the ADC of suspicious breast lesions in quantitative DWI of the breast was assessed in chapter 4. The magnitude of the ADC value varied with the choice of different b-values. However, the diagnostic performance was not affected by the choice of the b-values. These results imply, that to differentiate benign from malignant breast lesions, the choice of the b-values is not relevant. However, when the absolute ADC value of a lesion would be compared to reported ADC threshold values in literature, the b-values should be taken into account. In our computer simulations and phantom experiments, we found that in the simple situation of a single hypointensity or calcification against a homogeneous background, the hypointensity or calcification could be visualized on MR images with both a negative and a positive contrast by applying a projection dephaser in the slice select direction (chapter 5). If these results could be extrapolated to the clinical situation of a patient with microcalcifications in the breast, a challenging diagnostic problem could potentially be overcome. The results described in chapter 6 show that MRI-guided 14-gauge core-needle biopsy at 3T is a safe and effective method for breast biopsy in lesions that are occult on mammography and ultrasound. The 9.5 mm susceptibility artifact caused by the biopsy needle did not hamper the procedure. In case of a benign biopsy result, 6 month follow-up Breast MRI should be performed. The performance of proton resonance frequency shift (PRFS) – based MR thermometry in clinical practise can be hampered by local magnetic field disturbances in the breast due to respiration and cardiac motion. We found these field disturbances in the breast to amount to approximately 0.13 ppm due to regular respiration and 0.16 ppm due to maximum capacity respiration and due to cardiac motion