Project Summary:
Degenerative cervical myelopathy (DCM) is the most common cause of non-traumatic spinal cord injury in older adults (over 50 years) worldwide. This is particularly relevant to the growing aging population (age greater than 65 years) of Wisconsin, which is projected to increase from 18% to 24% of the total population over the next 20 years. The pathophysiology of DCM is characterized by chronic spinal cord compression and surgical decompression of the cervical spinal cord is the primary treatment for DCM. Surgical decision-making is primarily based on clinical evaluation and evidence of cervical spinal cord compression on conventional magnetic resonance imaging (MRI). The degree of spinal cord compression on conventional MRI, however, correlates poorly with neurological dysfunction in myelopathy, yet this remains a major driver behind the decision to offer surgery. Additionally, existing quantitative analyses of pre-surgical MRIs are weak predictors of post-surgical recovery of function. Due to limited diagnostic and prognostic tools, diagnostic delays (one-to-five years) are common in DCM, and over 30% of people undergoing spine surgery for DCM do not achieve a meaningful recovery of neurological function. There is an unmet need for an objective biomarker to better select patients for surgical intervention in DCM. Without this, patients with DCM will continue to endure delays in diagnosis and may undergo surgery despite limited potential for neurological recovery.
Radiomic analysis converts medical images such as MRIs into high-dimensional, mineable feature space. Extracted features assess regional heterogeneity by quantifying variations in signal intensity within an MRI and measuring the relationship between neighboring voxels. MR-based radiomic features are sensitive to subtle changes in microstructure and have been used to diagnose and prognosticate patients with other spinal cord pathologies such as multiple sclerosis. Radiomic assessment of the spinal cord show promise as an imaging biomarker in DCM, yet the diagnostic and prognostic accuracy of this tool has not been evaluated in a uniform imaging dataset that included comparative healthy control data. By deriving objective data from MRI textures, radiomics has the potential to advance existing quantitative analyses of conventional MRIs. Radiomic texture analyses of standard clinical MR images have high translational value since this approach does not require additional MR imaging time or new equipment to implement in the clinic.
The long-term goal of this project is to develop an imaging biomarker for DCM. The overall objective of this project is to identify MRI-based radiomic signatures that are associated with baseline neurological function and can predict post-surgical neurological function. AHW funding will provide valuable data for a larger grant application aimed at validating our findings and establishing radiomic MRI analysis in clinical practice.
Since DCM is common in older adults, the target population to be impacted by this proposal is men and women over the age 50 years. Radiomic MRI biomarkers are expected to be incorporated into the diagnostic workup for DCM to reduce delays in diagnosis as well as to create personalized risk assessments for DCM patients regarding progression of disease. This is expected to address health inequality in DCM where certain race groups are more likely to face delays in diagnosis of DCM leading to greater disability, dependence on supportive care and unemployment. Incorporating radiomic MRI data will also strengthen the shared decision-making process with the patient in the clinic. Patients will have a better understanding of how much they could expect to improve after surgery and surgeons could better identify patients who would have a meaningful recovery of neurological function after surgery. Together, the addition of a radiomic MRI biomarker will substantially advance the clinical management of DCM and reduce disability associated with the disease among the aging population of Wisconsin.