Optimization and implementation of emerging technologies to improve radiation therapy of head-and-neck cancer

External beam radiotherapy (RT) is considered a cornerstone in the treatment of head-and-neck cancer (HNC). The main goal of RT is to administer a sufficient dose to induce a cell killing effect in the targeted malignant cells, hereby maximizing tumor control, while limiting the dose to the healthy organs-at-risk (OAR), hereby minimizing radiation-induced toxicities. Enhanced geometric targeting has been crucial to reduce toxicities in photon beam RT of HNC, through technical developments such as radiation beam intensity-modulation and image-guided treatment delivery. The increased time demands of these more complex treatment approaches conflict with the under-provision of RT, however, such that improved time-efficiency of state-of-the-art photon beam RT of HNC is required. In addition, the dose distributions achievable with photon beam RT are ultimately limited by the governing exponential dose deposition profile, which advocates the implementation of alternative RT beam modalities.  The absence of exit dose in intensity-modulated proton therapy (IMPT), for instance, due to the finite range of protons, could allow for improved OAR sparing compared with photon beam RT and may therefore reduce toxicities for a subpopulation of the patients with HNC. The relatively immature technique of IMPT, however, still requires optimization of many technologic and physical aspects, such as the lateral penumbra and the workflow efficiency when using a range shifter (RS) to treat superficial target volumes. This dissertation focusses on the optimization and implementation of emerging technologies to improve the abovementioned aspects in photon and proton beam RT of HNC.

Implementation of emerging O-ring linac technology may increase the time-efficiency of image-guided volumetric intensity-modulated arc therapy (VMAT) with photon beams. The encapsulation of the moving parts, namely, allows for higher gantry rotation speeds than with conventional C-arm linacs. In chapter II of this thesis, a fast-rotating O-ring linac was clinically implemented and compared with a C-arm linac in terms of plan quality and delivery time. For a cohort of 30 patients, treatment plans for VMAT of HNC on both systems were optimized, plan delivery accuracy was verified and plan delivery times were measured. The results show that for VMAT of HNC, the fast-rotating O-ring linac at least maintains the plan quality of dual-arc VMAT on a C-arm linac while reducing the volumetric image acquisition and plan delivery time. The expected reduction in treatment time associated with the fast-rotating O-ring linac could increase clinical throughput of image-guided, highly conformal photon beam RT of HNC and could contribute to improved patient comfort in standard- or hyper-fractionated RT of HNC. In addition, the fast-rotating O-ring linac provides inherent availability of daily volumetric images, which could facilitate clinical deployment of adaptive RT of HNC.

IMPT of HNC may be optimized by using bolus RS, integrated within the patient immobilization. Such integrated device could reduce the spot size and the number  of  hardware  manipulations  –  with  collision  risk  –  compared  with conventional nozzle-mounted RS. In chapters III, IV and V of this thesis, the emerging manufacturing technology of 3D printing was implemented towards the realization of such device. Firstly, the normal tissue sparing with 3D printed, patient-specific bolus RS for IMPT of HNC was compared with nozzle-mounted RS using Monte Carlo based plan optimization. The results show clinically relevant toxicity reductions and thus motivate the development of patient-specific bolus RS for IMPT of HNC. Secondly, the suitability of 3D printing to achieve RS integrated patient immobilization was evaluated for a range of 3D printed materials and techniques. Dual-energy computed tomography based range shift predictions were verified with proton beam measurements, and mechanical stiffness measurements were performed before and after irradiation with a therapeutic dose. The results show a proof-of-concept for the use of 3D printed RS and immobilization in a clinical workflow. Thirdly, the feasibility of 3D printed immobilization (3DPrIm) was evaluated in a pilot study on patients with HNC, using weekly cone-beam computed tomography imaging and a patient questionnaire. The results show that 3DPrIm is feasible in terms of workflow, patient comfort and setup reproducibility. As such, with the implementation of 3D printing technology, important steps were taken towards a single, patient-specific device that achieves spot size reduction, hardware manipulation minimization and patient immobilization for IMPT of HNC. While the clinical implementation of such device is yet to be achieved, the enhanced toxicity reduction could increase the overall number of patients eligible for IMPT of HNC. Moreover, the potentially increased clinical throughput could facilitate the clinical evidence gathering for IMPT of HNC.