Functional Imaging Guided Parotid Gland Sparing Radiation Therapy using Diffusion-Weighted Magnetic Resonance Imaging

Katelyn Cahill

doi:10.25907/01074

Despite technological advancements, radiation-induced xerostomia (dry mouth) remains one of the most common side effects reported in Head and Neck Cancer (HNC) patients. This is due to the incidental dose that the parotid glands receive during Radiation Therapy (RT) and their location in the head and neck. Due to their proximity to the target volume, exposure to this dose is unavoidable. Xerostomia can impact every-day activities such as sharing a meal, laughing with friends or joining a conversation. This often leads to a decrease in quality of life (QoL), which is why efforts are being made to investigate the radiosensitivity of the parotid gland and its ability to regenerate and recover after radiation damage. Traditional normal tissue complication probability models of the parotid gland are being questioned, with the knowledge that it comprises of heterogeneous regions with differing functional importance. It’s been reported that the major salivary duct houses the stem/progenitor cells which are instrumental in the differentiation and replacement of saliva producing acinar cells. There is evidence that sparing this region of the parotid gland results in reduced xerostomia where dose-volume relationships have been proven. This provides a strong biological rationale to investigate imaging techniques such as functional magnetic resonance imaging (MRI) sequences to determine whether they can identify such regions. This thesis investigates the use of Diffusion-Weighted Magnetic Resonance Imaging (DW-MRI) to potentially identify functionally critical subvolumes within the parotid gland using apparent diffusion coefficient (ADC) measurements. It also introduces the concept of a biological at-risk volume (BRV) of the parotid gland and aims to evaluate the use of ADC percentile thresholds to be used in personalised BRV parotid gland sparing RT. To our knowledge, the studies presented in this thesis are the first to propose and implement this approach for parotid subvolume delineation. Chapter 1 provides an overview of RT in HNC and describes the evolution of RT techniques and patient outcomes such as radiation-induced xerostomia toxicity. The impact of such advancements is not as widespread as predicted and the prevalence of such adverse events remains high. Here, DW-MRI is introduced as a functional imaging sequence with the potential to identify the highly cellular area within the parotid gland which could correspond with the region pertinent to regeneration at a cellular level. The application of DW-MRI and the associated ADC in HNC is discussed, but no evidence suggests its use in the delineation of functionally critical subvolumes. Chapter 2 presents a scoping review of the literature and charts the delineation strategies used in studies that investigate parotid subvolume sparing RT. Key themes were identified and reviewed including dose-volume histogram objectives as well as subvolume dose-response relationships with xerostomia, QoL and salivary flow outcomes. The included studies have primarily used anatomical boundaries to delineate parotid gland subvolumes. During the course of this thesis, new literature has been published with a focus on sparing the major parotid duct with the aim to reduce dose to the stem cell region. DW-MRI was not used in any of these investigations, identifying the limited use of functional imaging to identify the critically functional region of the parotid gland. This has highlighted a gap in this area of research and warrants further investigation. Chapter 3 was instrumental in evaluating departmental MRI imaging protocols and identifying challenges and inconsistencies in the diagnostic MRIs available at the time. This retrospective analysis helped form recommendations and allowed for the development of RT specific image acquisition protocols for application during RT planning. This foundation was critical for multidisciplinary education and upskilling which facilitated the changes to imaging acquisition protocols. Importantly, the completion of this phase has resulted in protocols that use consistent parameters across all patients, which in turn, validates and improves the reliability of ADC measurements. Chapter 4 describes the implementation of a retrofit MRI-Simulator (MRI-Sim) for an existing MRI within the medical imaging department and was published in the Journal of Medical Radiation Sciences in 2023. Having the ability to position HNC patients in their immobilisation devices during MRI-Sim increases the accuracy of application during RT planning. The evaluation of this service involved reviewing the sequences used to acquire DW-MRI to ensure high image quality, high with minimal geometric distortion, a high signal-to-noise ratio and acceptable acquisition time. The implementation of this service was pivotal to the success of the remaining phases of research in this thesis. Chapters 5 to 7 comprise of three core studies that address the primary aims and research questions of this thesis. Without the work completed in the previous two chapters, the DW-MRI derived ADC measurements would not have been reliable throughout this work. Chapter 5 developed a method to automatically delineate a parotid gland BRV and was published in the British Journal of Radiology Open in 2025. This phase tested various ADC percentile thresholds to determine the process of delineating the resultant BRV. Its distribution was evaluated throughout the parotid gland to ascertain whether there was any relationship between BRV location and the superficial lobe of the parotid where the stem cell region is situated. This study informed the percentile values used in the following retrospective planning study. Chapter 6 presents the findings of a retrospective feasibility dosimetry study which compares the dosimetry quality and dose to organs of interest, particularly the parotid gland, between standard whole parotid gland sparing versus DW-MRI guided parotid gland sparing RT. This study applies the BRV developed in Chapter 5 as an additional avoidance region during dose optimisation, this work has been submitted and under review with the Radiation Oncology Journal. These results conclude that the 30th ADC percentile measurement is the most feasible option for parotid gland BRV sparing which was then applied in the following translational research. Chapter 7 reports on the acute adverse events and QoL outcomes for the Functional Imaging Guided Salivary Gland Sparing Radiation Therapy (FIGS-RT) pilot study. This study facilitated the translation of research through the implementation of previously developed and tested workflows and successfully implemented this technique into clinical practice. This study aimed to investigate whether reducing dose to the parotid gland BRV has a predictive impact on xerostomia outcomes. It also considers if a dose-relationship can be demonstrated between the mean dose to the parotid gland BRV and xerostomia related side effects and patient QoL using variables such as area under the curve and time to peak adverse event. This study is ongoing, and the long-term outcomes will be reviewed outside the scope of this thesis. Aligning with the language changes proposed by the European Society for Radiotherapy and Oncology (ESTRO) in November 2025, Chapter 7 embraces a more balanced, patient-friendly language when describing radiation-induced toxicities, organs at-risk and dose constraints. Instead, terminology such as side effects or adverse events, organs of interest and dose guidance will be adopted to influence positive and empathetic perceptions with a patient-centred focus in our communication across scientific and everyday health care settings. It was important to make this transition, with the aim to submit the research in Chapter 7 to a European journal. In summary, this thesis successfully applied DW-MRI derived ADC measurements to form a parotid gland BRV, allowing for a personalised approach to subvolume sparing of the parotid gland in HN RT. This work was translated into clinical practice where early indications suggest a potential functional-dose relationship between the contralateral BRV mean dose and acute xerostomia and demonstrates a strong relationship with QoL 3-months post-RT.

Functional Imaging Guided Parotid Gland Sparing Radiation Therapy using Diffusion-Weighted Magnetic Resonance Imaging

Files and links (1)

Abstract

Details

Metrics