|
Pre-treatment
prediction of brain tumor response to radiation therapy using high
diffusion-weighted MRI
Yael
Mardor, Roberto Spiegelmann, Yiftach Roth, Arie Orenstein, Stephan E. Maier,
Ouzi Nissim, Zvi Ram, Jacob Baram, Thomas Tichler, Raphael Pfeffer
Sheba
Medical Center, Ramat-Gan, Israel; Brigham and Womens Hospital, Harvard Medical
School, Boston, MA
Introduction.
Diffusion-weighted MRI (DWMRI) allows for non-invasive characterization of
tissues based on the diffusion of water molecules.
Tumor water diffusion is correlated with tumor cellularity.
Our previous studies demonstrated that high diffusion-weighted MRI (HDWMRI),
which is more sensitive to the slow water population than conventional DWMRI, is
more susceptible to detect early changes in human brain tumors post therapy than
conventional DWMRI.
DWMRI can distinguish solid viable tumor from cystic and necrotic regions.
We have previously demonstrated the feasibility of using HDWMRI to predict
treatment outcome prior to initiation of chemotherapy in an animal model.
In this work we apply the use of pre-treatment HDWMRI to the prediction of
response to radiotherapy in brain tumor patients.
Objectives.
To study the application of high diffusion-weighted MRI performed before
treatment to predict response to radiotherapy in brain tumor patients.
Materials
and Methods.
Eight patients with nine brain lesions were studied.
Four patients had gliomas (grades III-IV) and four patients had brain metastasis
(three breast and one prostate cancer).
All patients received conventional fractionated radiation therapy.
The glioma patients received fractions of 2 Gy per day to a total of 54-60
Gy.
The patients with metastases received fractions of 2-3 Gy per day to a total of
30-40 Gy.
Conventional and high diffusion-weighted MRI (up to b=4000 sec/mm2) were
acquired on a 0.5 T interventional MRI with the Line Scan DWMRI software.
Response to therapy was determined from changes in tumor volumes calculated from
contrast-enhanced T1-weighted MRI, acquired before and 50 days on average after
initiation of therapy.
Diffusion-weighted images were analyzed and a diffusion index, R, reflecting
tissue viability based on the water molecular diffusion, was calculated.
The correlation between pre-treatment values of R and response or lack of
response was studied.
Results.
The diffusion index, R, for the pre-treatment tumors was found to correlate
significantly with later tumor response or lack of response (r2=0.51,
p<0.03).
This correlation implies that tumors with low pre-treatment R values, indicating
high viability, will respond better to radiation therapy than tumors with high R
values, indicating necrosis.
Conclusions.
The correlation between the pre-treatment diffusion index and later tumor
response indicates that HDWMRI may be used prior to initiation of treatment, to
predict the outcome of certain anti-tumor therapies, thus enabling individual
optimization of treatment plan.
We are currently widening our investigations to include chemotherapy for primary
CNS lymphomas and single fraction radiosurgery in brain tumor patients.
Copyright © 2003 American Association for Cancer Research. All rights reserved.
Source: http://aacr03.agora.com/planner/displayabstract.asp?presentationid=1363
|
