Apparent diffusion coefficient histogram parameters distinguish paediatric brain tumours
Year: 2008
Session type: Poster / e-Poster / Silent Theatre session
UCL Institute of Child Health, London, UK
Abstract
Background
Accurate preoperative diagnosis is important in paediatric patients with CNS lesions, particularly in those with posterior fossa (PF) lesions. Our aim was to determine whether it is possible to differentiate the common posterior fossa tumour types using diffusion-weighted MRI to measure the apparent diffusion coefficient (ADC) histogram within the tumour volume. In addition we specifically looked at differentiating a rare tumour type atypical teratoid rhabdoid tumour (ATRT) from its embryological relative primitive neurectodermal tumours–medulloblastoma (PNET-MB).
Method
Total of 56 patients were included in the study (23 females and 33 males) with a histopathologically confirmed diagnosis subsequent to their preoperative imaging. Mean age was 6.1 yrs (0.1-15.8). MRI performed using Siemens Symphony or Avanto 1.5T system.
Results
There were significant differences between PNET-MB’s and astrocytomas confirming previous work and for the first time, using only ADC data, between ependymomas and astrocytomas (p<0.05), ependymomas and PNET-MB’s (p<0.05) and between astrocytomas and ATRT (p<0.05).
Discussion
Application of logistic regression (LR) analysis to discrimination of common posterior fossa tumours was successful with a sample of 32 individuals and 3 tumour groups. LR to discriminate ATRT from PNET was 100% successful using the variables of peak height and mean ADC, which alone were inadequate. The data sets are small but the incidence of these tumours is relatively rare, hence the ADC map should be collected routinely. Common paediatric posterior fossa tumours are discriminated 94% successfully. ADC histogram analysis has the potential to aid non invasive diagnosis and obviate the need for surgical biopsy.
Background
Accurate preoperative diagnosis is important in paediatric patients with CNS lesions, particularly in those with posterior fossa (PF) lesions. Our aim was to determine whether it is possible to differentiate the common posterior fossa tumour types using diffusion-weighted MRI to measure the apparent diffusion coefficient (ADC) histogram within the tumour volume. In addition we specifically looked at differentiating a rare tumour type atypical teratoid rhabdoid tumour (ATRT) from its embryological relative primitive neurectodermal tumours–medulloblastoma (PNET-MB).
Method
Total of 56 patients were included in the study (23 females and 33 males) with a histopathologically confirmed diagnosis subsequent to their preoperative imaging. Mean age was 6.1 yrs (0.1-15.8). MRI performed using Siemens Symphony or Avanto 1.5T system.
Results
There were significant differences between PNET-MB’s and astrocytomas confirming previous work and for the first time, using only ADC data, between ependymomas and astrocytomas (p<0.05), ependymomas and PNET-MB’s (p<0.05) and between astrocytomas and ATRT (p<0.05).
Discussion
Application of logistic regression (LR) analysis to discrimination of common posterior fossa tumours was successful with a sample of 32 individuals and 3 tumour groups. LR to discriminate ATRT from PNET was 100% successful using the variables of peak height and mean ADC, which alone were inadequate. The data sets are small but the incidence of these tumours is relatively rare, hence the ADC map should be collected routinely. Common paediatric posterior fossa tumours are discriminated 94% successfully. ADC histogram analysis has the potential to aid non invasive diagnosis and obviate the need for surgical biopsy.
