MRI Texture Analysis as an Early Imaging Biomarker for Progressive Cerebral Adrenoleukodystrophy
Guardado en:
| Publicado en: | ProQuest Dissertations and Theses (2025) |
|---|---|
| Autor principal: | |
| Publicado: |
ProQuest Dissertations & Theses
|
| Materias: | |
| Acceso en línea: | Citation/Abstract Full Text - PDF |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Resumen: | X-linked adrenoleukodystrophy (X-ALD) is a rare peroxisomal disorder. A mutation in the ABCD1 gene causes this X-linked recessive disorder. This mutation causes the production of a defective ALD protein that is essential for the β-oxidation of very long fatty acids (i.e. those with more than 22 carbons) in the peroxisome. As a result, there is an increased level of VLCFs in the serum, brain, adrenal tissue, and testes. The effects of VLCF manifest in various toxic ways throughout the body. Approximately 35-40% boys with genetically positive X-ALD develop the childhood cerebral form of ALD of which ~75% have a progressive subtype with demyelination leading to neurodegeneration and death without treatment. While newborn screening has improved early detection and follow-up, it cannot predict which individuals will develop the childhood cerebral form (cALD), nor distinguish those with cALD who have this devastating progressive subtype from those who have a non-progressive subtype that self-halts without intervention. To determine the phenotype and subtype, monitoring requires serial gadolinium-enhanced MRI scans every six months from age 2 to 12 to identify the progressive cALD subtype early enough for timely treatment (e.g. with hematopoietic stem cell transplantation (HSCT)) when it is still efficacious.  To enhance treatment efficacy and minimize gadolinium exposure from serial MRI scans, it is crucial to identify early imaging biomarkers. The widespread use of texture analysis in neurological disorders motivates us to detect changes in normal appearing white matter (NAWM), as observed in histology and advanced imaging techniques in cALD. We hypothesized that MRI texture analysis can detect progressive cALD before overt demyelination and may serve as an early imaging biomarker for the detection of cALD.  We conducted a retrospective study using T1 and T2-weighted MRI data from 12 progressor and 24 non-progressor cases collected across three sites. To test our hypothesis that texture analysis can detect subtle microstructural changes and potentially serve as an early imaging biomarker for identifying the progressive form of X-linked adrenoleukodystrophy (X-ALD), we designed two complementary studies with a retrospective dataset. The first study focused on NAWM to assess significant texture feature differences between two groups of X-ALD. The second study examined the lesion area to determine if there are any texture feature differences and to compare these findings with those from the first study. For the first study, region of interest (ROI) selection was made from NAWM. The ROI for this study included the parietal white matter, periatrial white matter, corpus callosum (genu, body), middle cerebellar peduncles, cerebral peduncles, thalamus, and centrum semiovale (upper and lower). Gray-level co-occurrence matrix (GLCM) texture features were extracted using the radiomics module in 3D Slicer. For the second study, the lesions served as the ROI, and we extracted the same GLCM texture features except from T2-weighted images. Statistically significant differences (p < 0.05) were observed in study 1 in the regions of lower centrum semiovale, anterior periventricular white matter, corpus callosum, and left cerebral peduncle in the NAWM of T1-weighted MRI. Additionally, significant differences were observed in the T2-weighted MRI of the parietal white matter. Statistical analysis was performed using the Mann-Whitney U test and false discovery rate correction was applied to control for multiple comparisons. The significant texture features were heterogeneity, homogeneity, and entropy for T1-weighted MRI.  In Study 2, we focused on T2-weighted lesions and analyzed the similar GLCM features within the lesion ROI (region of interest). This study also revealed statistically significant differences between groups in features such as heterogeneity, homogeneity, entropy, summary, and linearity. Across the two studies, we observed that some statistically significant texture feature differences emerged in NAWM and persisted in the lesion as well. These findings are consistent with the histological findings of NAWM/prelesion area. These changes may indicate the presence of dysmyelination or early changes of demyelination or gliosis, or hypomyelination too subtle to detect using conventional visual assessments. The findings support the potential benefit of texture analysis as an early imaging biomarker for progressive childhood cerebral ALD (cALD). Further validation with a larger prospective dataset and correlation with other quantitative imaging techniques will provide additional confidence as to the utility of this imaging biomarker.  |
|---|---|
| ISBN: | 9798293830480 |
| Fuente: | ProQuest Dissertations & Theses Global |