A Dissertation Presented to the Faculty of The University of Texas MD Anderson Cancer Center UTHealth Graduate School of Biomedical Sciences
in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
by Abdallah Sherif Radwan Mohamed, M.D., M.Sc.
Thesis Advisory Professor:
Clifton D. Fuller, M.D., Ph.D.
Thesis Committee Members:
Robert Bast, MD
Stephen Y. Lai, MD, PhD
Jihong Wang, PhD
Jason Stafford, PhD
Panayiotis Mavroidis, PhD
Presented in Public Defense & Examination in
Houston, Texas, and virtually via Teleconference, on
March 30th, 2022.
Thesis Abstract
The aim of the clinical component of this work was to determine whether the currently
available clinical imaging tools can be integrated with radiotherapy (RT) platforms for
monitoring and adaptation of radiation dose, prediction of tumor response and disease
outcomes, and characterization of patterns of failure and normal tissue toxicity in head and
neck cancer (HNC) patients with potentially curable tumors. In Aim 1, we showed that the
currently available clinical imaging modalities can be successfully used to adapt RT dose
based-on dynamic tumor response, predict oncologic disease outcomes, characterize RT-induced toxicity, and identify the patterns of disease failure. We used anatomical MRIs for the
RT dose adaptation purpose. Our findings showed that after proper standardization of the
immobilization and image acquisition techniques, we can achieve high geometric accuracy.
These images can then be used to monitor the shrinkage of tumors during RT and optimize the
clinical target volumes accordingly. Our results also showed that this MR-guided dose
adaptation technique has a dosimetric advantage over the standard of care and was
associated with a reduction in normal tissue doses that translated into a reduction of the odds
of long-term RT-induced toxicity.
In the second aim, we used quantitative MRIs to determine its benefit for prediction of
oncologic outcomes and characterization of RT-induced normal tissue toxicity. Our findings
showed that delta changes of apparent diffusion coefficient parameters derived from diffusion-weighted images at mid-RT can be used to predict local recurrence and recurrence free survival. We also showed that K
trans and Ve vascular parameters derived from dynamic contrast-enhanced MRIs can characterize the mandibular areas of osteoradionecrosis.
In the final clinical aim, we used CT images of recurrence and baseline CT planning
images to develop a methodology and workflow that involves the application of deformable
image registration software as a tool to standardize image co-registration in addition to granular
combined geometric- and dosimetric-based failure characterization to correctly attribute sites
and causes of locoregional failure. We then successfully applied this methodology to identify
the patterns of failure following postoperative and definitive IMRT in HNC patients. Using this
methodology, we showed that most recurrences occurred in the central high dose regions for
patients treated with definitive IMRT compared with mainly non-central high dose recurrences
after postoperative IMRT. We also correlated recurrences with pretreatment FDG-PET and
identified that most of the central high dose recurrences originated in an area that would be
covered by a 10-mm margin on the volume of 50% of the maximum FDG uptake.
In the translational component of this work, we integrated radiomic features derived
from pre-RT CT images with whole-genome measurements using TCGA and TCIA data. Our
results demonstrated a statistically significant associations between radiomic features
characterizing different tumor phenotypes and different genomic features. These findings
represent a promising potential towards non-invasively tract genomic changes in the tumor
during treatment and use this information to adapt treatment accordingly. In the final project of
this dissertation, we developed a high-throughput approach to identify effective systemic
agents against aggressive head and neck tumors with poor prognosis like anaplastic thyroid
cancer. We successfully identified three candidate drugs and performed extensive in vitro and
in vivo validation using orthotopic and PDX models. Among these drugs, HDAC inhibitor and
LBH-589 showed the most effective tumor growth inhibition that can be used in future clinical
trials.
This is a deposition of the Thesis version presented for Public Defense and Examination, and preliminarily approved by the Thesis Committee on 2022-03-30; a finalized version will deposited at the Texas Medical Center Digital Commons [https://digitalcommons.library.tmc.edu/utgsbs_dissertations/] upon formal Graduate School of Biomedical Sciences administrative approval, proofing/formatting, and doctoral degree conferral.
The presented Public Defense slide (2022-03-30) presentation is deposited as a PowerPoint and PDF version.
Funding
Longitudinal Spatial-Nonspatial Decision Support for Competing Outcomes in Head and Neck Cancer Therapy
Using Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) to Establish Objective Clinical Outcome Measures for Mandibular Osteoradionecrosis
National Institute of Dental and Craniofacial Research