Publication: Valor diagnóstico de las señales de susceptibilidad intratumoral en gliomas difusos del adulto
Authors
Tudela Martínez, José Ignacio
item.page.secondaryauthor
Escuela Internacional de Doctorado
item.page.director
Berná Mestre, Juan de Dios
Publisher
Universidad de Murcia
publication.page.editor
publication.page.department
DOI
item.page.type
info:eu-repo/semantics/doctoralThesis
Description
Abstract
Los gliomas difusos del adulto constituyen el grupo más frecuente de tumores primarios malignos del sistema nervioso central y se caracterizan por una marcada heterogeneidad clínica, radiológica y molecular. La quinta edición de la clasificación de tumores del SNC de la Organización Mundial de la Salud (OMS SNC 5, 2021) ha consolidado un modelo diagnóstico integrado histomolecular, en el que entidades como el astrocitoma isocitrato deshidrogenasa (IDH)-mutado, el oligodendroglioma IDH-mutado y 1p/19q co-delecionado y el glioblastoma IDH-wildtype se definen por alteraciones moleculares específicas. No obstante, el diagnóstico integrado continúa dependiendo de procedimientos invasivos para la obtención de tejido tumoral, lo que limita la caracterización biológica completa del tumor debido a la heterogeneidad intratumoral y al riesgo quirúrgico asociado.
En este contexto, las técnicas de resonancia magnética (RM) avanzada, y particularmente la imagen ponderada en susceptibilidad magnética (SWI), han emergido como herramientas prometedoras para el desarrollo de biomarcadores no invasivos. Las señales de susceptibilidad intratumoral (ITSS), descritas como estructuras hipointensas puntiformes o lineales en SWI, reflejan fenómenos como angiogénesis aberrante, microhemorragias y necrosis microscópica, estrechamente vinculados a la agresividad tumoral. La presente tesis se compone de dos estudios consecutivos. En el primer estudio, se realizó un análisis semicuantitativo de las ITSS en una cohorte retrospectiva de gliomas difusos del adulto, aplicando criterios diagnósticos actualizados según la OMS SNC 5. Se analizó su capacidad para predecir el grado tumoral (alto vs. bajo grado) mediante modelos de regresión logística y su correlación con parámetros histológicos, inmunohistoquímicos y moleculares. Los resultados confirmaron una asociación significativa entre mayor grado de ITSS y mayor agresividad tumoral, así como correlaciones consistentes con biomarcadores moleculares clave, especialmente con el estado IDH-wildtype y la presencia de deleción homocigota de los genes inhibidores de la quinasa dependiente de ciclina 2A y 2B (CDKN2A/B).
El segundo estudió se centró específicamente en glioblastomas IDH-wildtype y se desarrolló un modelo cuantitativo basado en análisis fractal de las ITSS. Se calcularon parámetros de dimensión fractal (FD) y lacunaridad (LAC) para caracterizar la complejidad geométrica y la heterogeneidad espacial de las ITSS. Posteriormente, se construyeron modelos logísticos univariantes y multivariantes para la predicción no invasiva de la deleción homocigota de CDKN2A/B. Los resultados mostraron que mayores valores de FD se asociaron significativamente con la presencia de dicha alteración molecular, sugiriendo que la complejidad estructural de las ITSS puede reflejar desregulación del ciclo celular y mayor agresividad biológica.
En conjunto, esta tesis demuestra que las ITSS, tanto mediante evaluación semicuantitativa como a través de modelos fractales avanzados, constituyen biomarcadores de imagen estructurales con potencial valor radiomolecular en la era del diagnóstico integrado. Estos hallazgos refuerzan el papel de la SWI como herramienta complementaria en la caracterización no invasiva de los gliomas difusos del adulto y abren nuevas líneas de investigación orientadas a integrar biomarcadores de imagen en modelos predictivos clínicamente aplicables dentro del paradigma de la radiogenómica.
Adult-type diffuse gliomas constitute the most common group of primary malignant tumors of the central nervous system and are characterized by marked clinical, radiological, and molecular heterogeneity. The fifth edition of the World Health Organization Classification of Central Nervous System Tumors (WHO CNS 5, 2021) has consolidated an integrated histomolecular diagnostic model, in which entities such as isocitrate dehydrogenase (IDH)-mutant astrocytoma, IDH- mutant and 1p/19q-codeleted oligodendroglioma, and IDH-wildtype glioblastoma are defined by specific molecular alterations. Nevertheless, integrated diagnosis continues to rely on invasive procedures for obtaining tumor tissue, which limits comprehensive biological characterization due to intratumoral heterogeneity and the associated surgical risks. In this context, advanced magnetic resonance imaging (MRI) techniques, particularly susceptibility-weighted imaging (SWI), have emerged as promising tools for the development of non-invasive imaging biomarkers. Intratumoral susceptibility signals (ITSS), described as punctate or linear hypointense structures on SWI, reflect phenomena such as aberrant angiogenesis, microhemorrhages, and microscopic necrosis, all closely linked to tumor aggressiveness. This thesis is structured into two consecutive studies. In the first one, the semiquantitative assessment of ITSS was evaluated in a retrospective cohort of adult diffuse gliomas using the updated diagnostic criteria according to WHO CNS 5. Their ability to predict tumor grade (high vs. low grade) was analyzed using logistic regression models, along with correlations with histological, immunohistochemical, and molecular parameters. The results confirmed a significant association between higher ITSS grades and increased tumor aggressiveness, as well as consistent correlations with key molecular biomarkers, particularly IDH-wildtype status and the presence of homozygous deletion of the cyclin-dependent kinase inhibitor 2A and 2B genes (CDKN2A/B). The second study focused specifically on IDH-wildtype glioblastomas, and a quantitative model based on fractal analysis of ITSS was developed. Fractal metrics including fractal dimension (FD) and lacunarity (LAC) were calculated to characterize the geometric complexity and spatial heterogeneity of ITSS. Univariate and multivariate logistic regression models were constructed for the non-invasive prediction of CDKN2A/B homozygous deletion. The results showed that higher FD values were significantly associated with the presence of this molecular alteration, suggesting that the structural complexity of ITSS may reflect cell cycle dysregulation and increased biological aggressiveness. Overall, this thesis demonstrates that ITSS, assessed both semiquantitatively and through advanced fractal modelling, constitute structural imaging biomarkers with potential radiomolecular value in the era of integrated diagnosis. These findings reinforce the role of SWI as a complementary tool for the non-invasive characterization of adult diffuse gliomas and open new avenues of research aimed at integrating imaging biomarkers into clinically applicable predictive models within the paradigm of radiogenomics.
Adult-type diffuse gliomas constitute the most common group of primary malignant tumors of the central nervous system and are characterized by marked clinical, radiological, and molecular heterogeneity. The fifth edition of the World Health Organization Classification of Central Nervous System Tumors (WHO CNS 5, 2021) has consolidated an integrated histomolecular diagnostic model, in which entities such as isocitrate dehydrogenase (IDH)-mutant astrocytoma, IDH- mutant and 1p/19q-codeleted oligodendroglioma, and IDH-wildtype glioblastoma are defined by specific molecular alterations. Nevertheless, integrated diagnosis continues to rely on invasive procedures for obtaining tumor tissue, which limits comprehensive biological characterization due to intratumoral heterogeneity and the associated surgical risks. In this context, advanced magnetic resonance imaging (MRI) techniques, particularly susceptibility-weighted imaging (SWI), have emerged as promising tools for the development of non-invasive imaging biomarkers. Intratumoral susceptibility signals (ITSS), described as punctate or linear hypointense structures on SWI, reflect phenomena such as aberrant angiogenesis, microhemorrhages, and microscopic necrosis, all closely linked to tumor aggressiveness. This thesis is structured into two consecutive studies. In the first one, the semiquantitative assessment of ITSS was evaluated in a retrospective cohort of adult diffuse gliomas using the updated diagnostic criteria according to WHO CNS 5. Their ability to predict tumor grade (high vs. low grade) was analyzed using logistic regression models, along with correlations with histological, immunohistochemical, and molecular parameters. The results confirmed a significant association between higher ITSS grades and increased tumor aggressiveness, as well as consistent correlations with key molecular biomarkers, particularly IDH-wildtype status and the presence of homozygous deletion of the cyclin-dependent kinase inhibitor 2A and 2B genes (CDKN2A/B). The second study focused specifically on IDH-wildtype glioblastomas, and a quantitative model based on fractal analysis of ITSS was developed. Fractal metrics including fractal dimension (FD) and lacunarity (LAC) were calculated to characterize the geometric complexity and spatial heterogeneity of ITSS. Univariate and multivariate logistic regression models were constructed for the non-invasive prediction of CDKN2A/B homozygous deletion. The results showed that higher FD values were significantly associated with the presence of this molecular alteration, suggesting that the structural complexity of ITSS may reflect cell cycle dysregulation and increased biological aggressiveness. Overall, this thesis demonstrates that ITSS, assessed both semiquantitatively and through advanced fractal modelling, constitute structural imaging biomarkers with potential radiomolecular value in the era of integrated diagnosis. These findings reinforce the role of SWI as a complementary tool for the non-invasive characterization of adult diffuse gliomas and open new avenues of research aimed at integrating imaging biomarkers into clinically applicable predictive models within the paradigm of radiogenomics.
publication.page.subject
Citation
item.page.embargo
Collections
Ir a Estadísticas
Este ítem está sujeto a una licencia Creative Commons. http://creativecommons.org/licenses/by-nc-nd/4.0/