Glioblastoma multiforme (GBM) is the most aggressive and most common primary brain tumor in adults accounting for 50-60% of primary brain tumors. The prognosis for patients with GBM remains poor and treatment is mainly palliative with a mean survival time of less than one year. Radiotherapy is used extensively in the management of glioblastoma either alone or in combination with surgery and/or chemotherapy. However, this tumor is one of the most resistant tumors to radiotherapy thus limiting the benefit of this form of treatment. <p>Studies have shown that malignant tumors have a high content of glutathione an antioxidant responsible for protecting the cells against damage from free radicals (mainly superoxide, hydroxyl and hydrogen peroxide). It is well established that glutathione, by neutralizing these free radicals plays a major role in radioresistance. Glioblastoma has relatively high levels of glutathione. In this study, by reducing the glutathione content of glioblastoma in a rat model, we were able to investigate the effect of this reduction in enhancing the effect of radiotherapy as a form of treatment for glioblastoma multiforme in a rat model. <p>By injecting L-Buthionine-SR-Sulfoximine (BSO) in to the tumor tissue, the glutathione content of the tumor was reduced by about 70% of its initial value. When administered into the tumors 2 hours prior to radiotherapy the animals so treated had a significantly longer median survival time compared with animals that received radiotherapy alone.

In recent years, there has been much improvement in radiation therapy delivery systems used in the treatment of cancer; however, in order to fully exploit this enhancement, the computational methodology associated with radiation therapy must improve as well. It is important to accurately determine where the radiation is depositing its energy within the patient. The treatment should allow for the maximal dose at the tumor site, while minimal radiation dose to the surrounding health tissue and structures. In the Computational Reactor and Medical Physics Group here at Georgia Tech, a heterogeneous coarse-mesh transport method (COMET) has been developed for neutron transport to analyze whole-core criticality. COMET decomposes a large, heterogeneous global problem into a set of small fixed source local problems. Response functions, or rather detailed solutions, are obtained for each unique local problem. These response functions are all precomputed and stored in a library. The solution to the global solution is then bound by a linear superposition of the local problems. In this project, COMET is applied for the first time to the transport of photons in human tissues. The parameter of interest in this case is the amount of energy (dose) deposition in tissue. To determine the strengths and weaknesses of the current system, it is important to construct benchmark problems for comparison. This project will encompass a number of benchmarks. The first will involve modeling a simple two-dimensional water phantom. A second benchmark problem involves the use of a heterogeneous phantom composed of different tissues. A third benchmark problem will involve transport through slabs of aluminum, water, and lung tissue. A last, more clinically relevant benchmark problem will involve using the data from a CT scan. For each of these cases the results from COMET will be compared to the computational results obtained from EGSnrc, a Monte Carlo particle transport code. In this study, it was found that generally the results using COMET were comparable with those obtained from the Monte Carlo solutions of EGSnrc. The COMET results were also typically found thousands of times faster than the reference solution.

Equispaced beam arrangements are typically used for IMRT plans. This beam arrangement provides adequate dose coverage to the target while sparing dose to other structures. However, an equispaced beam arrangement may not provide the best dose coverage to the target while sparing dose to the other structures. Beam angle optimization attempts to optimize the beam directions to produce a better IMRT plan; this is achieved by increasing dose to the target while minimizing dose to the remaining structures. Most methods of beam angle optimization attempt to optimize the beam angles and the beam intensity profiles to find an optimal set of beam angles. This thesis attempts to optimize the beam angles without determining the beam intensity profiles. An MCNP simulation is run to score the beam directions; the simulation is run as an adjoint problem to reduce simulation time, with the target as the source and the detectors scoring the dose for the gantry angles of the beam. Then, an optimization algorithm is run to select a set of beam angles for an optimized IMRT plan. The optimized IMRT plan is compared to an equispaced IMRT plan on a commercial treatment planning system to determine if this method of beam angle optimization is better than using an equispaced beam arrangement. The results of this thesis indicate that the coupling of an MCNP simulation for scoring with an optimization algorithm to select beam angles will produce a better IMRT plan than an equispaced IMRT plan. Three different geometries were used and for all geometries, the optimized IMRT plan had a higher average dose to the target while maintaining or increasing dose sparing to the critical structure and normal tissue.

Accurate tumor tracking remains as a major challenge in radiation therapy. Large margins are added to the clinical target volume (CTV) to ensure the treatment of tumor in presence of patient setup uncertainty and that caused by intra-motion. Fiducial seeds and calypso markers are commonly implanted into the disease sites to further reduce the dose delivery error due to tumor motion. For more accurate dose delivery and improved patient comfort, the use of radioactive tracers in positron emission tomography (PET) as non-invasive tumor markers has been proposed - a concept called emission-guided radiation therapy (EGRT). Instead of using images obtained from a stand-alone PET scanner for treatment guidance, we mount a positron imaging system on a radiation therapy machine. Such an EGRT system is able to track the tumor in real time based on the lines of response (LOR) of the tumor positron events, and perform radiation therapy simultaneously. In this work, we illustrate the EGRT concept using computer simulations and propose a typical treatment scheme. EGRT's advantage on increased dose delivery accuracy is demonstrated using a pancreas tumor case and a lung tumor case without the setup margin and motion margin. The emission process is simulated by Geant4 Application for Tomographic Emission package and Linac dose delivery is simulated using a voxel-based Monte Carlo algorithm. The tumor tracking error can be controlled within 2 mm which indicates margins can be significantly reduced. The dose distributions show that the proposed EGRT can accurately deliver the prescribed dose to the CTV with much less margins. Although still in a preliminary research stage, EGRT has the potential to substantially reduce tumor location uncertainties and to greatly increase the performance of current radiation therapy.

Die fraktionierte stereotaktisch geführte Radiotherapie stellt eine Möglichkeit der Behandlung intrakranieller Tumoren dar, bei denen eine chirurgische Resektion nicht oder nicht vollständig möglich ist. Ziel dieser Arbeit war der Vergleich der Dosisverteilung von Bestrahlungsplänen, die mittels fluenzmodulierter (IMRT) und 3D konformaler stereotaktisch geführter Bestrahlungstechnik (conformal beam) für Hypophysenadenome und petroclivale Meningeome erstellt wurden. Auf Basis des CT-Datensatzes eines Alderson-Phantoms erfolgten zunächst die systematische Untersuchung der IMRT Eingabeparameter und die Definition von Dosis-Volumen-Constraints für beide Tumorentitäten. Im Anschluss wurden die gewonnenen Erkenntnisse bei der fluenzmodulierten Bestrahlungsplanung von 10 realen Patienten, 5 je Tumorentität, angewendet. Zusätzlich wurde für jeden dieser Patienten ein 3D konformaler Bestrahlungsplan erstellt. Die Evaluation der Bestrahlungspläne erfolgte durch verschiedene quantitativer Parameter unter Einbeziehung der subjektiven Beurteilung der Dosisverteilung. Bei beiden Tumorentitäten lieferte in je 4 von 5 Fällen die IMRT einen Vorteil gegenüber der conformal beam Bestrahlung, der vor allem auf der besseren Erfassung, Dosishomogenität und höheren Median-Dosis im bestrahlten Zielvolumen (PTV) beruhte. Komplex geformte Zielvolumina profitierten dabei besonders von der Fluenzmodulation. Dem stand jedoch in einigen Fälle eine stärke Belastung der Risikoorgane (OAR) mit einer höheren Dosis und die generell höhere Anzahl notwendiger Monitoreinheiten gegenüber. Zuletzt wurde auf Basis von MatLAB eine Software entwickelt, welche die Verifikation von stereotaktisch geführten IMRT-Bestrahlungen und die Analyse von Winston-Lutz-Tests ermöglicht.

Diss. (sammanfattning) Stockholm : Karolinska institutet, 2010.

published_or_final_version / Dentistry / Doctoral / Doctor of Philosophy

Σκοπός της παρούσας εργασίας είναι η μελέτη με την βοήθεια της τεχνικής Monte Carlo των φασμάτων εκτός πεδίου του γραμμικού επιταχυντή SL- 18 της ELEKTA. Η μελέτη έδειξε συσχέτιση μεταξύ δόσης και απόστασης από των άξονα της δέσμης. / The aim of this study was to investigate the energy spectrum outside the field limits of therapeutic high energy photon beams. Monte Carlo simulations were used in order to determine the dose and to confirm or to try to re-estimate the Energy correction factors used in off-field in-vivo dosimetry. The ELEKTA-SL18 medical accelerator was simulated for 6MV and 15MV, using the EGSnrc code. The simulation includes the regular jaws and the MLC. The output of each simulation was a square scoring plane at SSD 100cm. The energy spectrum, the mean energy, the energy fluence and other parameters were studied for annular areas centred on the Z axis, and for 1cm2 rectangular areas centred on both X and Y axis. These regions were selected every 1cm inside and outside the reference (10x10) cm2 field of the primary beam. The spectra and all the aforementioned parameters were found to be in relation to the position, and their comparison revealed differences, that exceeded the statistical error, between areas that had the same distance from the center but were located on different axes. These differences were more important for the lower energy (6MV), as the contribution from leakage radiation is relatively higher. Their comparison served to study the influence of the spectral differences on the measurements of this energy-dependent dosimetric system outside the treatment field. The ELEKTA SL-18 LINAC was simulated for photon beams of 6MV and 15MV with and without MLCs. The photon energies and the dose to the out-field areas close to treatment field are considerable and this should be taken into account when radiosensitive organs are close to the field limits. This could be more important to complicated IMRT treatments where the treatment time is altered.

- / -

An ultra-short echo time (UTE) magnetic resonance imaging (MRI) sequence was used together with other MRI sequences to evaluate the possibility of segmenting air, soft tissues and bone. Three patients were imaged with the UTE sequence and other sequences as well as with computed tomography (CT). An algorithm using Gaussian mixture models was developed and applied to the problem of segmenting the MR images. A similar algorithm was developed and used to generate an artificial CT image from the MR data. The images of the first patient were used as training data for the algorithms and the images of the other two patients were used for validation. It was found that less than 20 percent of the volume inside the head was misclassified and that the root mean square error of the artificial CT image was less than 420 Hounsfield units. Finally a volunteer was imaged in the same way but with an additional UTE sequence with a larger flip angle. The results suggested that the additional image may improve segmentation further. / En sekevens för bildgivande magnetresonans (MRI) med ultrakort ekotid (UTE) användes tillsammans med andra MRI-sekvenser till att utvärdera möjligheten att segmentera luft, mjukvävnad och ben. Bilder togs av tre patienter med UTE-sekvensen och med övriga sekvenser samt med datortomografi (CT). En algoritm baserad på en blanding av normalfördelningar utvecklades och tillämpades på MR-segmenteringsproblemet.En likande algoritm utvecklades och användes till att skapa en konstgjord CT-bild utifrån MR-bilderna.Bilderna tagna av den första patienten användes till att träna algoritmerna medan bilderna av de två andra patienterna användes för validering. Mindre än 20 procent av volymen inuti huvudet felklassificerades och det kvadratiska medelvärdet av avvikelserna i den konstgjorda CT-bilden var mindre än 420 hounsfieldenheter. Slutligen togs bilder av en frivillig på samma sätt men med ytterligare en UTE-sekvens med en större flippvinkel. Resultatet antyder att den nya bilden kan bidra till en förbättrad segmentering.