Author ORCID Identifier
https://orcid.org/0009-0001-9300-4857
Date Available
11-1-2025
Year of Publication
2025
Document Type
Doctoral Dissertation
Degree Name
Doctor of Philosophy (PhD)
College
Medicine
Department/School/Program
Radiation Medicine
Faculty
Damodar Pokhrel
Faculty
Ellis Lee Johnson
Abstract
As technology advances, more and more cancer patients are benefiting from the use of stereotactic radiosurgery and radiotherapy (SRS/SRT) to treat their disease. One such technology, Varian’s HyperArc high-definition radiosurgery module, a tool to help facilitate the treatment of brain metastases on a medical linear accelerator (LINAC), has been demonstrated to achieve comparable low dose spillage and target conformity compared to the current gold standard for intracranial malignancies, the Elekta Gamma Knife. Furthermore, due to HyperArc’s single-isocenter multiple lesion approach, the treatment can be delivered in a significantly shorter amount of time. Unfortunately, the manual process of creating these stereotactic treatment plans is a meticulous, time-consuming task that requires the expertise of an experienced treatment planner. Often consisting of a simulation to treatment time of 1 to 2 weeks, as opposed to the maximum simulation to treatment time of 6 hours for Gamma Knife. Additionally, the quality of these manually generated plans can be heavily dependent on the planner. During this time the tumors are continuing to grow inside the patient’s brain. This results in a suboptimal plan where the tumor’s size and location planned on, has the possibility of changing prior to the delivery of the generated plan. To succinctly state the issues observed: the quality of radiation therapy treatment should not be dependent on where the treatment is done or who is doing the treatment. Furthermore, when a diagnosis and course of action is established, treatments should be prepared and delivered as fast as possible.
This dissertation develops, demonstrates, and clinically implements software and methodologies to automate and standardize the treatment planning process, facilitating a shorter simulation to treatment time including a “same-day” planning process for the treatment of single-isocenter single and multiple brain lesions (SISL/SIML) through HyperArc SRS and SRT. The primary tools used for this dissertation were Varian’s HyperArc, knowledge-based planning module RapidPlan, Multi-Criteria Optimization (MCO), Eclipse Scripting Application Programming Interface, and Varian’s Medical Affairs’ free and open-source Plan Scorecards. The provided tools were manipulated and refactored into a single, streamlined user interface to allow automated treatment planning with minimal user interaction while still maintaining sufficient autonomy over critical decisions for an effective plan quality. Achieving a decrease in treatment planning time from 2 to 4 hours for manual SRS/SRT planning to less than 20 minutes with the automated planning process for SISL/SIML brain SRS/SRT plans.
Additionally, to demonstrate the robust capabilities of the tools and processes developed throughout this research, the treatment planning process for recurrent head and neck cancers was automated. Recurrent head and neck cancers are notoriously difficult cases to plan for due to the involved anatomical landscape where many organs have already received a full course of previous dose, resulting in a difficult optimization problem of sufficiently providing dose to effectively kill the tumor cells, while attempting to minimize dose to nearby organs to improve patient quality-of-life and minimize adverse side-effects. To further highlight the possibilities enabled with automated treatment planning, a tumor dose escalation study was conducted for select recurrent head and neck patients, demonstrating the ability of the software to aggressively minimize dose to nearby organs while sufficiently improving the effect of the radiation on the tumor cells.
With the intention of decreasing the simulation to treatment time through the automation and standardization of the treatment planning process, the results of this dissertation demonstrate high-quality clinically acceptable SRS/SRT plans for both brain metastases and recurrent head and neck cancers that can be generated automatically in under 20 minutes with robust methodology and software that is readily available and intuitive to use. With fast plan generation, quality assurance, and delivery, the simulation to treatment time can be reduced from 1 to 2 weeks to an estimated 4 to 6 hours, comparable to that of Gamma Knife SRS, but on a LINAC.
Digital Object Identifier (DOI)
https://doi.org/10.13023/etd.2025.97
Funding Information
This study was partly supported by Varian Medical Systems Grant #3048115893 from 2023 to 2025.
Recommended Citation
McCarthy, Shane, "Development and Clinical Validation of Automated Same-Day Stereotactic Treatment Procedures for Multiple Brain Metastases and Recurrent Head and Neck Cancers" (2025). Theses and Dissertations--Radiation Medicine. 7.
https://uknowledge.uky.edu/radmed_etds/7
