Glioblastoma (GBM) spreading occurs through several potential pathways: perivascular space, cerebrospinal fluid, brain parenchyma and white matter tracts (WMTs). In recent studies the hypothesis of spreading along WMTs appears the most valuable and explain the relationship with survival. GBM locating in brain regions with higher WMTs density have more potential spreading routes and shorter overall survival (OS). In a recent study higher tract density index (TDI) values are associated with poorer prognosis, likely reflecting an increased potential for tumor dissemination in regions with greater white matter fiber density, where more anatomical pathways for spread are available. Nonetheless the role of WMTs in GBM spreading remains only partially accepted. In this study, we focused on gliomas with distant relapses, which represent a paradigmatic and informative model for investigating the spreading mechanisms according to structural connectivity. The objective of this thesis is to examine the relationship between distant glioblastoma recurrences and the structural organization of the brain. The study aims to determine whether the primary tumor and its distant relapses tend to occur along the same white matter tracts, as opposed to being randomly distributed within the cerebral parenchyma. The result will finally support, or not, the hypothesis of cancer spreading along WMTs route. The study included patients diagnosed with GBM who developed distant recurrences identifiable on follow-up MRI scans. For each patient, MRI data, demographic characteristics, clinical-prognostic variable, overall survival and the progression free survival. Two synthetic lesion distribution models were created for comparison: one based on randomly redistributing GBM lesions between different patients, and the other one based on generating randomly distributed glioma-like lesions. Within each subject (both the main GBM cohort and synthetic patients), atlas-based metrics were computed, including the tract involvement index (TII), quantifying the degree of co-involvement of the same white matter tracts by the primary and corresponding relapse lesions, and the disconnectome involvement index (DII), assessing the spatial overlap between the recurrence foci and the primary lesion’s disconnectome. Additionally, for each lesion, the tract density index (TDI) was calculated, defined as the average number of white matters streamlines per voxel in the lesion mask. Forty-two cases of gliomas with distant relapses were included in the study. TII values were significantly different between the groups, as confirmed by one-way ANOVA model (F=4.98, p=0.008). Post-hoc t-test confirmed that TII was higher in gliomas with distant relapses compared to redistributed glioma lesions (p-value=0.006). Gliomas with distant recurrences did not exhibit significantly different DII values compared to randomly distributed spherical lesions (one tailed z-test=1.18, p=0.12; 2=1.363, p=0.243). Additionally, a negative correlation was observed between time to recurrence and the overall survival (R=-0.31, p=0.03), as well as a marginally significant correlation between time to recurrence and the primary lesion-relapse distance (R=-0.29, p=0.051). In GBM with distant recurrences, higher TII and DII values suggest that distant relapses are more likely to develop along the same white matter tracts affected by the primary tumor, rather than appearing in a spatially random manner. This finding confirms the hypothesis that white matter tracts may serve as preferential routes for tumor migration. Moreover, the observed correlations between earlier recurrence, short overall survival and greater lesion separation reinforce the notion that more biologically aggressive tumors exhibit faster relapses and more extensive dissemination
Il glioblastoma (GBM) può diffondersi attraverso diverse potenziali vie: spazio perivascolare, liquido cefalorachidiano, parenchima cerebrale e tratti di sostanza bianca (WMTs). In studi recenti, l'ipotesi della diffusione lungo i WMT appare la più valida. I GBM localizzati in aree cerebrali con una maggiore densità di WMT presentano più potenziali vie di diffusione e una sopravvivenza globale (OS) più breve. In uno studio recente, valori elevati di tract density index (TDI) sono stati associati a una prognosi peggiore, riflettendo una maggiore capacità di disseminazione in aree con maggiore densità di WMT, dove sono presenti più vie anatomiche per la diffusione. Tuttavia, il ruolo dei WMT nella diffusione del GBM rimane solo parzialmente accettato. In questo studio, ci siamo concentrati sui gliomi con recidive a distanza, che rappresentano un modello paradigmatico per indagare i meccanismi di diffusione in base alla connettività strutturale. L'obiettivo di questa tesi è valutare la relazione tra le recidive di glioblastoma a distanza e l'organizzazione strutturale del cervello. Lo studio mira a determinare se il tumore primario e le sue recidive a distanza tendano a manifestarsi lungo gli stessi tratti di sostanza bianca, anziché essere distribuiti casualmente all'interno del parenchima cerebrale. Lo studio ha incluso pazienti con diagnosi di GBM che hanno sviluppato recidive a distanza identificabili tramite RM di follow-up. Per ciascun paziente sono stati raccolti i dati radiologici, le caratteristiche demografiche, le variabili cliniche, la OS e la progression free survival. Sono stati creati due modelli sintetici di distribuzione delle lesioni come controllo: uno basato sulla ridistribuzione casuale delle recidive tra diversi pazienti e l'altro basato sulla generazione di lesioni sferiche simil-glioma distribuite casualmente. Per ciascun soggetto (sia della coorte principale di GBM che per i pazienti sintetici), sono state calcolati parametri, tra cui il tract involvement index (TII), che quantifica il grado di coinvolgimento degli stessi tratti di sostanza bianca da parte delle lesioni primarie e delle corrispondenti recidive, e il disconnectome involvement index (DII), che valuta la sovrapposizione spaziale tra la recidiva e il disconnettoma della lesione primaria. Inoltre, per ciascuna lesione è stato calcolato il tract density index (TDI), definito come il numero medio di fasci di sostanza bianca per voxel nella maschera della lesione. Quarantadue casi sono stati inclusi nello studio. I valori di TII erano significativamente diversi tra i gruppi, come confermato dal modello ANOVA a un fattore (F=4,98, p=0,008). Il t-test post-hoc ha confermato che il TII era più elevato nei gliomi con recidive a distanza rispetto alle recidive ridistribuite (valore p=0,006). I gliomi con recidive a distanza non hanno mostrato valori di DII significativamente diversi rispetto alle lesioni sferiche distribuite casualmente (z-test a una coda=1,18, p=0,12; chi quadro=1,363, p=0,243). Inoltre, è stata osservata una correlazione negativa tra il tempo diagnosi-recidiva OS (R=-0,31, p=0,03), nonché una correlazione marginalmente significativa tra il tempo diagnosi-recidiva e la distanza lesione primaria-recidiva (R=-0,29, p=0,051). Nel GBM con recidive a distanza, valori di TII e DII più elevati suggeriscono che le recidive abbiano maggiori probabilità di svilupparsi lungo gli stessi tratti di sostanza bianca interessati dal tumore primario, piuttosto che manifestarsi in modo spazialmente casuale. Questo risultato conferma l'ipotesi che i tratti di sostanza bianca possano fungere da vie preferenziali per la migrazione del tumore. Inoltre, le correlazioni osservate tra recidiva precoce, breve OS e maggiore distanza tra le lesioni lesioni rafforzano l'idea che i tumori biologicamente più aggressivi presentino recidive più rapide e una disseminazione più estesa.
Spatial distribution of distant glioblastoma relapses in relation to the structural connectome
ROSA, ANTONIO
2024/2025
Abstract
Glioblastoma (GBM) spreading occurs through several potential pathways: perivascular space, cerebrospinal fluid, brain parenchyma and white matter tracts (WMTs). In recent studies the hypothesis of spreading along WMTs appears the most valuable and explain the relationship with survival. GBM locating in brain regions with higher WMTs density have more potential spreading routes and shorter overall survival (OS). In a recent study higher tract density index (TDI) values are associated with poorer prognosis, likely reflecting an increased potential for tumor dissemination in regions with greater white matter fiber density, where more anatomical pathways for spread are available. Nonetheless the role of WMTs in GBM spreading remains only partially accepted. In this study, we focused on gliomas with distant relapses, which represent a paradigmatic and informative model for investigating the spreading mechanisms according to structural connectivity. The objective of this thesis is to examine the relationship between distant glioblastoma recurrences and the structural organization of the brain. The study aims to determine whether the primary tumor and its distant relapses tend to occur along the same white matter tracts, as opposed to being randomly distributed within the cerebral parenchyma. The result will finally support, or not, the hypothesis of cancer spreading along WMTs route. The study included patients diagnosed with GBM who developed distant recurrences identifiable on follow-up MRI scans. For each patient, MRI data, demographic characteristics, clinical-prognostic variable, overall survival and the progression free survival. Two synthetic lesion distribution models were created for comparison: one based on randomly redistributing GBM lesions between different patients, and the other one based on generating randomly distributed glioma-like lesions. Within each subject (both the main GBM cohort and synthetic patients), atlas-based metrics were computed, including the tract involvement index (TII), quantifying the degree of co-involvement of the same white matter tracts by the primary and corresponding relapse lesions, and the disconnectome involvement index (DII), assessing the spatial overlap between the recurrence foci and the primary lesion’s disconnectome. Additionally, for each lesion, the tract density index (TDI) was calculated, defined as the average number of white matters streamlines per voxel in the lesion mask. Forty-two cases of gliomas with distant relapses were included in the study. TII values were significantly different between the groups, as confirmed by one-way ANOVA model (F=4.98, p=0.008). Post-hoc t-test confirmed that TII was higher in gliomas with distant relapses compared to redistributed glioma lesions (p-value=0.006). Gliomas with distant recurrences did not exhibit significantly different DII values compared to randomly distributed spherical lesions (one tailed z-test=1.18, p=0.12; 2=1.363, p=0.243). Additionally, a negative correlation was observed between time to recurrence and the overall survival (R=-0.31, p=0.03), as well as a marginally significant correlation between time to recurrence and the primary lesion-relapse distance (R=-0.29, p=0.051). In GBM with distant recurrences, higher TII and DII values suggest that distant relapses are more likely to develop along the same white matter tracts affected by the primary tumor, rather than appearing in a spatially random manner. This finding confirms the hypothesis that white matter tracts may serve as preferential routes for tumor migration. Moreover, the observed correlations between earlier recurrence, short overall survival and greater lesion separation reinforce the notion that more biologically aggressive tumors exhibit faster relapses and more extensive dissemination| File | Dimensione | Formato | |
|---|---|---|---|
|
Rosa_Antonio.pdf
accesso aperto
Dimensione
1.77 MB
Formato
Adobe PDF
|
1.77 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/86974