Diffusion tensor imaging discriminates between glioblastoma and cerebral metastases in vivo
Year: 2008
Session type: Poster / e-Poster / Silent Theatre session
1St. George's, Univeristy of London, London, UK, 2UCL Institute of Child Health, London, UK
Abstract
Aims
To investigate the ability of diffusion tensor imaging (DTI) to discriminate between and predict histological diagnosis of the 2 most prevalent brain tumours; cerebral metastasis and glioblastoma.
Method
The study was approved for ethics and all patients gave their informed consent. In a prospective study, patients with a radiologically proven brain tumour underwent DTI prior to diagnosis and definitive treatment. The 28 patients who, were proven to have a glioblastoma or metastasis on histological examination were included in this study. Following definition of regions of interest, DTI metrics (mean diffusivity (MD) and fractional anisotropy (FA)) were calculated for the tumour volume and the surrounding region of peritumoural oedema. These metrics were then subjected to logistic regression to investigate their ability to discriminate between glioblastomas and cerebral metastases.
Results
The logistic regression analysis correctly predicted histological diagnosis of glioblastoma in 15/16 (93.8%), and metastasis in 11/12 (91.7%) of cases. MD was significantly higher and FA was relatively lower within the oedema surrounding metastases than within the oedema around glioblastomas. MD was significantly higher within the tumour volume of the glioblastomas.
Conclusion
In this study we demonstrate that when DTI metrics from the tumour volume and surrounding peritumoural oedema are studied in combination the effect is to be able to correctly distinguish between glioblastoma and cerebral metastases in the majority of cases. Development of quantitative imaging parameters such as those obtained from DTI may help to obviate the need for invasive biopsy in the future.
Aims
To investigate the ability of diffusion tensor imaging (DTI) to discriminate between and predict histological diagnosis of the 2 most prevalent brain tumours; cerebral metastasis and glioblastoma.
Method
The study was approved for ethics and all patients gave their informed consent. In a prospective study, patients with a radiologically proven brain tumour underwent DTI prior to diagnosis and definitive treatment. The 28 patients who, were proven to have a glioblastoma or metastasis on histological examination were included in this study. Following definition of regions of interest, DTI metrics (mean diffusivity (MD) and fractional anisotropy (FA)) were calculated for the tumour volume and the surrounding region of peritumoural oedema. These metrics were then subjected to logistic regression to investigate their ability to discriminate between glioblastomas and cerebral metastases.
Results
The logistic regression analysis correctly predicted histological diagnosis of glioblastoma in 15/16 (93.8%), and metastasis in 11/12 (91.7%) of cases. MD was significantly higher and FA was relatively lower within the oedema surrounding metastases than within the oedema around glioblastomas. MD was significantly higher within the tumour volume of the glioblastomas.
Conclusion
In this study we demonstrate that when DTI metrics from the tumour volume and surrounding peritumoural oedema are studied in combination the effect is to be able to correctly distinguish between glioblastoma and cerebral metastases in the majority of cases. Development of quantitative imaging parameters such as those obtained from DTI may help to obviate the need for invasive biopsy in the future.
