Low-grade (WHO II) and anaplastic (WHO III) gliomas: differences in morphology and MRI signal intensities.
Eur Radiol
; 23(10): 2846-53, 2013 Oct.
Article
em En
| MEDLINE
| ID: mdl-23686293
OBJECTIVE: To evaluate the diagnostic potential of a multi-factor analysis of morphometric parameters and signal characteristics of brain tumours and peritumoural areas for distinguishing WHO-grade II and III gliomas at magnetic resonance imaging (MRI). METHODS: MR examinations of 108 patients with histologically proven World Health Organization (WHO) grade II and III gliomas were included. Morphological criteria and MR signal characteristics were evaluated. The data were subjected to a multifactorial logistic regression analysis to differentiate between grade II and grade III gliomas. The discriminatory power was assessed by receiver operating characteristic (ROC). RESULTS: Logistic regression analysis showed that WHO grade II and III can be distinguished based on contrast enhancement, cortical involvement, margin of the enhancing lesion and maximum diameter (width and length) of the peritumoural area (the so-called tumour infiltration zone). With the final model of logistic regression analysis and with the cut-off value ≥ 0.377, WHO grade III glioma is predicted with a sensitivity of 71.0 % and a specificity of 80.4 %. CONCLUSION: Measurement of maximum diameter of peritumoural area, contrast enhancement as well as cortical involvement and the margin of the contrast-enhancing lesion can be used easily in clinical routine to adequately distinguish WHO grade II from grade III gliomas. KEY POINTS: ⢠MRI offers new information concerning WHO-grade II and III gliomas. ⢠The differentiation between such tumour grades is important for therapeutic decisions. ⢠We assessed differences in enhancement, cortical involvement, margins and peritumoural appearances. ⢠WHO grade III gliomas can be predicted with reasonable sensitivity and specificity.
Texto completo:
1
Coleções:
01-internacional
Base de dados:
MEDLINE
Assunto principal:
Algoritmos
/
Neoplasias Encefálicas
/
Imageamento por Ressonância Magnética
/
Interpretação de Imagem Assistida por Computador
/
Glioma
Tipo de estudo:
Diagnostic_studies
/
Prognostic_studies
Limite:
Adult
/
Aged
/
Aged80
/
Female
/
Humans
/
Male
/
Middle aged
Idioma:
En
Ano de publicação:
2013
Tipo de documento:
Article