|Title: ||Adaptation and verification of: intensity modulated radiotherapy & volumetric modulated arc therapy|
|Other Titles: ||Adaptatie en verificatie van: Intensiteit gemoduleerde radiotherapie en volume gemoduleerde boogtherapie|
|Authors: ||Crijns, Wouter; S0160674|
|Issue Date: ||30-Apr-2015 |
|Abstract: ||Prostate cancer is the most frequently diagnosed cancer in males in Belgium. In 2011, approximately 8300 of the 34000 newly diagnosed cancers in males were prostate cancer. This high incidence is in agreement with the incidence in other industrialized countries. Of those newly diagnosed cases, about fifty cases per year are treated with external beam radiotherapy at the department of radiation oncology of UZ Leuven. This work studies techniques to improve the quality of these external beam treatments.|
Current state of the art external beam radiotherapy techniques, such as Intensity Modulated Radiotherapy (IMRT) and Volumetric Modulated Arc Therapy (VMAT) administer the intended homogenous dose very tight (conformal) to the target, using steep dose gradients. Moreover, IMRT and VMAT can use the steep dose gradients to deposit a prescribed non-homogeneous target dose, i.e. dose painting. As a consequence IMRT and VMAT create new treatment options such as a simultaneous integrated boost for suspicious intra-prostatic lesions allowing dose distributions tailored to the tumor.
A first challenge of IMRT and VMAT is that it relies on complex optimization and sophisticated dose calculation algorithms to result in a unique, non-standard, treatment plan for each patient. Therefore, it is preferred to verify the quality of such a unique patient plan prior to the treatment. Additionally, the optimization and dose calculation algorithms demand finding the correct beam parameter settings using a set of dose measurements of standard fields representing the IMRT and VMAT treatment techniques. To offer a solution for both the patient specific quality assurance (QA) and the measurement set of standard fields, the first goal of this project is to design a practical methodology for multidimensional high resolution dose measurements.
A second challenge of IMRT and VMAT optimization is the dependence on a patient model containing: the delineations of the various organs and target regions; and the physical properties necessary to calculate the dose. A multi-modal imaging (CT, MR, PET, etc.) can provide such a patient model. Such an extensive model remains only a snap-shot of the patient’s anatomical dynamics, limiting the validity of the model in time. As a consequence the optimized IMRT or VMAT plan is only temporarily valid. Therefore, the second goal of the project is the design of a methodology to update the initial modulated treatment plan to the current anatomical situation, i.e. treatment adaptation.
1 Multidimensional high resolution dose measurements
Gafchromic film, with its high resolution and low perturbation of the measurement environment, is the preferred dosimeter for IMRT and VMAT. Additionally, this type of films shows to be convenient in clinical practice through its room light handling and off—the—shelf read out with a flatbed scanner.
Despite the favorable characteristics of these films, some important challenges block a swift implementation. (1) The nonlinear dose response lacks a proper modeling (and physical interpretation). (2) A systematic read out error restricts the size of the film. (3) Inter- and intra-batch variations impact the (zero dose) film response. Using two calibration films a robust calibration methodology simultaneously models the nonlinear dose response and the systematic read out error (challenge 1 and 2). The non-linear model provides a physical interpretation of the dose response. The two calibration films are scanned prior to scanning of measurement films. This methodology calibrates the scanner-film-transmission system for each measurement accounting for film response variations (challenge 3). The procedure is designed using phantom measurements and simulations of different calibration approaches. Next, it is validated with a different set of phantom measurements. Finally, a few patient examples illustrate the clinical applicability.
A remaining challenge is (4) the decreased signal-to-noise-ratio for low doses. This challenge hampers dose measurements at the distal edge of treatment fields as well as the dose out-of-field. Therefore, the signal-to-noise-ratio for measurements in these regions is increased using an increased dose (factor 10 or 100). This approach brings the in-field measurement far outside the calibration range invalidating this part of the measurement. A companding technique which combines different factors (e.g. 1, 10, and 100) however, allows measurements of the complete extend of the treatment beam. Such a method is useful for the measurement of the set of dose measurements of standard beams, which are required to find the correct beam parameter settings of the optimization and dose calculation algorithms.
2 Treatment adaptation
IMRT and VMAT contribute to individualized medicine by providing unique patient specific treatment plans. However, these treatment plans are only valid for the unique patient anatomy at the time of treatment imaging. Fortunately, various imaging modalities can provide an update of this anatomical situation, all with their own temporal and anatomical limitations. Consequently the IMRT and VMAT treatment plans can be updated to the current anatomical situation: i.e. treatment adaptation.
This work investigates the feasibility of a geometric and a dosimetric IMRT and VMAT adaptation based on the limited input of the positions of four fiducial markers implanted in the prostate and a more recent (CB)CT. To keep the adaptation procedure transparent only basic treatment plan corrections are applied such as: point dose calculations parallel to the treatment beam; and transformations, rotations and isotropic scaling perpendicular to the treatment beam.
The ultimate goal is a real-time dosimetric adaptation for VMAT. This goal was approached in three phases of increasing complexity. The first stage introduces the proposed limited input and basic corrections for adaptation of IMRT plans in response to inter-fraction anatomical variations. The IMRT adaptation is validated in a simulation study and illustrated on a single patient case. The patient case reveals the need for a quality assurance (QA) tool accompanying the adaption procedure.
In the second stage such a QA tool is provided together with an inter-fraction VMAT adaptation. The QA tool compared the adapted plan to a database of plans to decide on the normality of the adapted plan. Finally in the last stage, an intra-fraction VMAT adaptation is evaluated. Due to the interplay between modulated radiotherapy and the real-time target motion only actual dose measurements are able to integrate the composite dose. Therefore, the dose measurement techniques of the first part are applied to evaluate the intra-fraction VMAT adaptation.
In conclusion a robust calibration methodology for EBT film allows dose measurements with 2% accuracy. The methodology contributed to the daily clinical practice as a QA tool for more than 800 patient specific plans for both established and new treatment techniques, and for new treatment units.
The IMRT and VMAT adaptations applied in this dissertation showed better performance than our current clinical practice. This illustrates the powerful combination of a limited input (four fiducial points + a CT-image) and basic corrections (aperture translations, rotations, and scaling + point dose calculations) in adapting IMRT and VMAT in response to inter- and intra-fraction anatomical variations.
|Publication status: ||published|
|KU Leuven publication type: ||TH|
|Appears in Collections:||Laboratory of Experimental Radiotherapy|
ESAT - PSI, Processing Speech and Images