|
|
Metabolic
imaging of low-grade gliomas with three-dimensional magnetic resonance
spectroscopy
Pirzkall
A, Nelson SJ, McKnight TR, Takahashi MM, Li X, Graves EE, Verhey LJ,
Wara WW, Larson DA, Sneed PK
Department of Radiation Oncology, University of California, San
Francisco, School of Medicine, San Francisco, CA 94143, USA.
pirzkall@radonc17.ucsf.edu
Purpose. The role of radiotherapy (RT) seems established for patients
with low-grade gliomas with poor prognostic factors.
Three-dimensional
(3D) magnetic resonance spectroscopy imaging (MRSI) has been reported
to be of value in defining the extent of glioma infiltration.
We
performed a study examining the impact MRSI would have on the routine
addition of 2-3-cm margins around MRI T2-weighted hyperintensity to
generate the treatment planning clinical target volume (CTV) for
low-grade gliomas.
Methods
and Materials. Twenty patients with supratentorial gliomas WHO Grade
II (7 astrocytomas, 6 oligoastrocytomas, 7 oligodendrogliomas)
underwent MRI and MRSI before surgery.
The MRI was contoured manually;
the regions of interest included T2 hyperintensity and, if present,
regions of contrast enhancement on T1-weighted images.
The 3D-MRSI
peak parameters for choline and N-acetyl-aspartate, acquired
voxel-by-voxel, were categorized using a choline/N-acetyl-aspartate
index (CNI), a tool for quantitative assessment of tissue metabolite
levels, with CNI 2 being the lowest value corresponding to tumor.
CNI
data were aligned to MRI and displayed as 3D contours.
The
relationship between the anatomic and metabolic information on tumor
extent was assessed by comparing the CNI contours and other
MRSI-derived metabolites to the MRI T2 volume.
Results.
The limitations in the size of the region "excited" meant
that MRSI could be used to evaluate only a median 68% of the T2 volume
(range 38-100%), leaving the volume T2c.
The CNI 2 volume (median 29
cm(3), range 10-73) was contained totally within the T2c in 55% of
patients.
In the remaining patients, the volume of CNI 2 extending
beyond the T2c was quite small (median 2.3 cm(3), range 1.4-5.2), but
was not distributed uniformly about the T2c, extending up to 22 mm
beyond it.
Two patients demonstrated small regions of contrast
enhancement corresponding to the regions of highest CNI.
Other
metabolites, such as creatine and lactate, seem useful for determining
less and more radioresistant areas, respectively.
Conclusion.
Metabolically active tumor, as detected by MRSI, is restricted mainly
to the T2 hyperintensity in low-grade gliomas, but can extend outside
it in a limited and nonuniform fashion up to 2 cm.
Therefore, a CTV
including T2 and areas of CNI extension beyond the T2 hyperintensity
would result in a reduction in the size and a change in the shape of
the standard clinical target volumes generated by adding uniform
margins of 2-3 cm to the T2 hyperintensity.
PMID: 12128127 [PubMed - indexed for MEDLINE]
|
