Advanced Imaging of Disease: Fetal Alcohol Spectrum Disorders and Duchenne Muscular Dystrophy

Abstract

Advanced imaging techniques are being increasingly utilized as quantitative biomarkers and outcome parameters to identify and monitor disease and evaluate new treatments. Validation of these techniques in animal models is imperative to reliably interpret results and apply them clinically in humans. Studies reported here utilized computed tomography (CT) and magnetic resonance imaging (MRI) to characterize the sheep model of fetal alcohol spectrum disorders (FASD) and the golden retriever (GRMD) model of Duchenne muscular dystrophy (DMD). Objective biomarkers are necessary to help identify some children with (FASD) who do not demonstrate the facial dysmorphology criterion for diagnosis. The sheep model of first trimester binge alcohol exposure was used to study the utility of CT and MRI to distinguish disease. Based on CT assessment, total skull bone volume was significantly more sensitive than cranial circumference in identifying 6-month-old lambs exposed to binge alcohol. Poor nutrition may also contribute to adverse developmental outcomes of prenatal alcohol exposure and choline supplementation has shown benefit in rodent models. Magnetic resonance imaging (MRI) was used to study the brains of 6-month-old lambs from alcohol-exposed pregnant sheep supplemented with and without choline throughout pregnancy. Total brain volume in lambs was reduced regardless of choline supplementation. Duchenne muscular dystrophy (DMD) is an X-linked disease caused by mutations in the dystrophin gene. Therapies for DMD must be assessed using objective biomarkers such as MRI to document potential benefit. The lack of studies correlating MRI indices and histopathologic lesions has limited interpretation. The pectineus muscle from GRMD dogs was used to register and correlate histology with MRI using a novel registration technique, compare histology segmentation between Aperio and a custom algorithm for automated histomorphometry, and compare metrics between 4.7T and 3T scanners. We found that histology and MR images co-registered but metrics were poorly correlated. Aperio and the custom algorithm had similar results and with optimization, it could be applied to the current research. Only T1 values were correlated between MRI scanners but neither T1 values, nor any of the 3T metrics correlated with histologic segmentations. Multiple imaging modalities outlined in this dissertation show the usefulness of objective quantitative techniques that could be adapted and applied to any disease entity.