MRI-Pathology Correlation for Image Analytics-Based Treatment Outcome Assessment and Margin Planning in Rectal Cancers
Abstract
Research Idea and Application: Up to 10,000 rectal cancer patients are annually subjected to an unneeded, morbid total mesorectal excision despite responding well to neo-adjuvant therapy, and over 25,000 rectal cancer sufferers have quality-of-life issues such as anastomotic leakage, pelvic sepsis, wound infection, and fecal incontinence due to morbid surgical procedures. Pre-surgical chemoradiation (to reduce tumor burden) results in appearance changes on subsequent magnetic resonance imaging (MRI) exams that obfuscate expert identification of residual disease. A majority of patients thus undergo radical excision surgery due to difficulties in assessing how well the tumor has responded to treatment in vivo. Further, these patients suffer from significant chance of local recurrence despite aggressive treatment. There is thus a pressing clinical need to accurately assess response to treatment in vivo to inform the next steps (surgical guidance, intra-operative radiation) in rectal cancers. The PI has previously shown that computational tools may be able to discriminate benign treatment-related changes on MRI from residual cancer. These tools extract image representations that accentuate differences between residual cancer and benign radiation-related treatment effects. The goal of our study is to develop novel computerized tools that utilize standard-of-care post-treatment MRIs to provide clinically actionable information to better (a) guide surgical and follow-on therapy, and (b) identify which patients will benefit from radical surgery and which could safely avoid it, in rectal cancers. This approach will uniquely jointly leverage radiology and pathology information to spatially label regions of residual cancer and treatment effects on post-treatment MRI. This curated learning set will ensure highly accurate machine learning tools for creating the GPS and risk score. Contributions: Successful completion of our project will result in: -- A computerized "surgical GPS" (CaST) that will impact all 40,000 rectal cancer patients diagnosed annually (over 1000 of whom are military Veterans), by providing a navigation map that guides the surgeon with an estimate of surgical margins (extent of residual disease after chemoradiation) and impact on unresectable surrounding structures (for directed intra-operative radiation or follow-up chemotherapy). This will significantly improve survival and morbidity rates in rectal cancer patients. -- A computerized risk assessment score (CoRRA) to identify patients who will most benefit from follow-on resection surgery (i.e., have tumor remaining after chemoradiation) vs. those who may benefit from less radical treatment (i.e., have minimal or no tumor after chemoradiation). This will directly impact up to 10,000 patients who are annually subjected to an unneeded total mesorectal excision despite responding well to neo-adjuvant therapy, by significantly reducing unnecessary radical procedures and concomitant quality-of-life issues. Career Goals: The Principal Investigator (PI), Dr. Satish Viswanath, is a Research Assistant Professor in Biomedical Engineering at Case Western Reserve University, who will utilize this award to develop an independent research program in computational image analytics for image-guided cancer interventions, with particular focus on colorectal cancers. This will include creation of novel radiomics and radiology-pathology co-registration tools for interventional guidance and treatment response characterization via MRI. Having been previously recommended for pre- and post-doctoral DoD/CDMRP fellowships, the PI is highly suited for this award, and will leverage: -- Unique clinical expertise in colorectal cancer: Specifically for this project, the PI will be directly working and collaborating with three different renowned colorectal surgeons. This will provide him clear, clinical perspective in understanding the nuances of rectal cancer treatment a
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 31, 2017
- Source ID
- W81XWH1610329
Entities
People
- Satish E Viswanath
Organizations
- Case Western Reserve University
- United States Army