UCLA

Stereotactic body radiation therapy (SBRT) has become a standard-of-care option for localized prostate cancer, utilizing large fractionated dose in five or fewer fractions. Furthermore, SBRT may also be recommended for prostate patients who undergo radical prostatectomy as there is a high risk of biochemical recurrence. Despite improvement in treatment efficiency, SBRT toxicity remains a significant challenge for prostate and prostate bed patients. The development of MR-guided radiation therapy systems (MRgRT) has allowed the acquisition of high soft tissue contrast MRI for treatment planning, adaptation, and monitoring, which may be used to treatment toxicity. This dissertation sought to utilize a 0.35T MRgRT system to develop new tools and features to help manage prostate and prostate bed SBRT treatment and toxicity management. First, an air cavity electron density correction method was proposed for MR-guided online adaptive radiation therapy (MRgART) for more accurate dose calculation and better daily dose constraint management. MRgART air cavity electron density correction was shown to be unnecessary and clinically insignificant on low-field systems as SBRT plans were dosimetrically robust to air cavity variation. Next, a half-Fourier acquisition single-shot turbo spin echo (HASTE) MRI sequence was proposed to visualize the urethra with the aim to enable urethra dose constraining radiation therapy to mitigate genitourinary (GU) toxicity. HASTE demonstrated high urethra visibility with the potential to be used for treatment planning. Furthermore, HASTE was utilized to analyze urethral inter-fractional geometric and dosimetric variations. Urethra inter-fractional motion and anatomical changes were shown to result in daily urethral constraint failure. Therefore, on-board urethra imaging and treatment adaption may be warranted to protect the urethra. Lastly, a method to identify dosimetric predictors of GU toxicity, utilizing fraction-wise dose volume histogram metrics, based on a planning dose registration surrogate method, of urinary structures/substructures, was proposed. Planning dose registration was shown to be an accurate surrogate to the daily delivered dose for toxicity regression. Moreover, a logistic regression model using the identified dosimetric predictors was able to predict GU toxicity with an accuracy/sensitivity/specificity of 0.79/ 0.67/0.82. Fraction-wise urinary structure and substructure dosimetric predictors can be used to provide stable dosimetric metrics to guide treatment planning.