Overview
Primary intra-axial brain tumors account for approximately two thirds of all brain neoplasms, whereas the remaining one third is made up of metastases. As a group, gliomas are the most common brain tumors and include astrocytomas, oligodendrogliomas, ependymomas, and choroid plexus tumors.[1] Astrocytomas account for approximately 80% of all gliomas and are the most common supratentorial tumor in all age groups.[2] See images of brain astrocytomas below.
Grade II astrocytoma in a 27-year-old woman. Nonenhanced CT scan shows a heterogeneous, ill-defined, hypoattenuating area in the right temporal lobe.
Grade III astrocytoma in a 33-year-old woman. Nonenhanced studies (right) show a mixed-attenuation lesion (solid and cystic areas) in the right parietal lobe with adjacent vasogenic edema. After contrast enhancement (left), the solid component is enhancing.
Grade IV astrocytoma in a 73-year-old man. Top row (left to right), Nonenhanced CT images and fluid-attenuated inversion recovery (FLAIR) MRI. Bottom row, Axial nonenhanced and enhanced and coronal enhanced T1-weighted MRIs. CT demonstrates an inhomogeneous area of abnormal attenuation in the right temporal lobe that extends to the parietal region, with surrounding edema and mass effect. Enhanced MRI demonstrates heterogeneous enhancement, extensive vasogenic edema, and mass effect. Note the ependymal and subependymal enhancement involving the adjacent lateral ventricle and enhancement of the adjacent dura; this finding is consistent with spread. Classification of astrocytomas
The following World Health Organization (WHO) classification system for astrocytomas is the most widely accepted system:
- WHO grade II astrocytoma - Diffuse astrocytomas, including fibrillary astrocytomas, protoplasmic astrocytomas, gemistocytic astrocytomas
- WHO grade III astrocytoma - Anaplastic astrocytomas
- WHO grade IV astrocytoma - Glioblastomas, including giant-cell glioblastomas and gliosarcomas
- WHO grade II astrocytoma - Pleomorphic xanthoastrocytoma
- WHO grade II astrocytoma - Choroid glioma
- Pilocytic astrocytomas
- Subependymal giant-cell astrocytomas
Characteristics of astrocytomas
Astrocytomas are often divided into circumscribed or infiltrating tumors. Pilocytic astrocytomas and subependymal giant-cell astrocytomas are in the circumscribed group because they tend to respect anatomic boundaries and because they do not invade. Grade II, III, and IV astrocytomas are infiltrating because of their tendency to insinuate and invade. Tumor cells are often found distant from the imaged mass.
Pleomorphic xanthoastrocytomas occupy an intermediate position; although they are well circumscribed and slow growing, malignant progression may occur.
Imaging of brain tumors
Brain-tumor imaging has dramatically progressed over the past few decades with the development and refinement of computed tomography (CT) scanning, magnetic resonance imaging (MRI), positron emission tomography (PET) scanning, and advanced MR sequences.
Linscott et al compared pilomyxoid astrocytomas and pilocytic astrocytomas to determine whether they had imaging characteristics that would help in distinguishing between them, since the histopathology of pilomyxoid astrocytomas is distinct from that of pilocytic astrocytomas. CT and MRI scans, pathology reports, and clinical information from 21 patients with pathology-confirmed pilomyxoid astrocytomas were retrospectively reviewed. The investigators found that almost 50% of the tumors were located outside the typical hypothalamic/chiasmatic region and that intratumoral hemorrhage occurred in 25% of patients.[3]
Preferred examination
At present, contrast-enhanced MRI is the imaging modality of choice. The large amount of streak artifact in the posterior fossa that can be encountered with CT scanning does not affect MRI. The sensitivity of MRI studies is 82-100%, and the specificity is 81-100%. The excellent intrinsic contrast of conventional MRI makes it a sensitive study. The addition of contrast material and additional sequences can substantially improve the specificity. Additional techniques include MR spectroscopy (MRS), which allows clinicians to characterize the chemical composition of the mass by determining the presence and/or alteration of components such as lactate, N -acetylaspartate (NAA), choline (Cho), and myo-inositol (Ins). See the images below.
Low-grade astrocytoma in a 52-year-old woman. Top row (left to right), Nonenhanced CT scan, fluid-attenuated inversion recovery (FLAIR) MRI, and diffusion-weighted MRI. Bottom row (left to right), Single-voxel spectroscopic image showing with region of interest, spectrum, and perfusion image. Images show an ill-defined, hypoattenuating lesion in the right centrum semiovale extending to the left side through the body of the corpus callosum. The lesion has increased signal intensity on the FLAIR image. The true nature of the lesion cannot be easily established on anatomic imaging. Magnetic resonance spectroscopy (MRS) shows mild elevation of the choline (Cho) peak in relation to the N-acetylaspartate (NAA) peak with an inverted lactate peak. The findings are compatible with a low-grade neoplasm. Two investigational sequences are often helpful in difficult cases, though the resultant images should be interpreted with caution. The first is perfusion-weighted imaging improves characterization of the tumor and aids in equivocal cases when other causes of signal abnormality are suggested, such as with demyelinating lesions, infarcts, and abscesses. The second is diffusion-tensor imaging, which can demonstrate the relationship of the tumor to white matter tracts.
Although MRI has distinct advantages over CT scanning, contrast-enhanced CT scanning is still used at many centers as the imaging modality for the evaluation of intra-axial mass lesions. The sensitivity of contrast-enhanced CT scanning is 65-100%, and the specificity is 72-100%. Positive aspects of CT scanning include relatively short scanning times, decreased cost, and an open environment for patients with claustrophobia.[4, 5]
Limitations of techniques
MRI and CT scanning can depict the gross morphologic characteristics of the tumor, its relationship with adjacent tissue, and certain aspects of the chemical composition of the tumor (with MRS). Perfusion-weighted and diffusion-tensor images must be interpreted with caution.
Tumor margins are often difficult to determine with accuracy. Studies have demonstrated that the extent of tumoral involvement in grades I-III astrocytomas is underestimated when current conventional imaging is used. Although MRS has been useful, it also causes underestimation of the tumor burden. Tumor cells have been demonstrated well beyond the margin of any imaging abnormality.
Although imaging is instrumental in diagnosing the tumor and in evaluating the extent of disease or recurrence, only biopsy helps in determining the grade of tumor.
Computed Tomography
The appearance of astrocytomas on CT scans partly depends on their grade. Low-grade astrocytomas typically appear as homogeneous areas of decreased attenuation. They are relatively well circumscribed and 20% have associated calcification. Although low-grade tumors usually do not enhance, rare tumors demonstrate minimal enhancement.
Specific low-grade tumors have imaging characteristics that can increase the specificity of CT scanning. Pilocytic astrocytomas often appear as a cystic lesion with an eccentric mural nodule that strongly enhances after the administration of contrast agent. Subependymal giant-cell astrocytomas are typically near the foramen of Monro and usually occur in patients with tuberous sclerosis. Pleomorphic xanthoastrocytomas are typically supratentorial, cortically based masses with strong heterogeneous enhancement. The adjacent dura and meninges often enhance, creating the dural-tail appearance.
Grade III astrocytomas appear more heterogeneous. Edema is often appreciated, calcification is rare, and the enhancement pattern is usually more pronounced.
Grade IV astrocytomas are even more heterogeneous than tumors of other grades on CT scans, and they almost always enhance strongly. Hemorrhage and necrosis are common, but calcification is not. Extensive edema and mass effect are usually appreciated. This grade often involves both hemispheres by spreading by means of the corpus callosum or commissures.
With CT scanning, the scarcity of edema and mass effect with low-grade lesions may make the true extent of pathology difficult to ascertain. Furthermore, small lesions may not be visible if contrast material is not used.
See images of brain astrocytomas below.
Grade II astrocytoma in a 27-year-old woman. Nonenhanced CT scan shows a heterogeneous, ill-defined, hypoattenuating area in the right temporal lobe.
Grade III astrocytoma in a 71-year-old man. Contrast-enhanced (left) and nonenhanced (right) images show a cystic lesion with thick walls in the left parietal lobe, with thick rim enhancement on the enhanced image. Moderate surrounding vasogenic edema causes mass effect on the atrium of the left lateral ventricle.
Grade IV astrocytoma in a 73-year-old man. Top row (left to right), Nonenhanced CT images and fluid-attenuated inversion recovery (FLAIR) MRI. Bottom row, Axial nonenhanced and enhanced and coronal enhanced T1-weighted MRIs. CT demonstrates an inhomogeneous area of abnormal attenuation in the right temporal lobe that extends to the parietal region, with surrounding edema and mass effect. Enhanced MRI demonstrates heterogeneous enhancement, extensive vasogenic edema, and mass effect. Note the ependymal and subependymal enhancement involving the adjacent lateral ventricle and enhancement of the adjacent dura; this finding is consistent with spread.
Recurrent grade IV astrocytoma in the region of the right caudate and putamen in a 76-year-old man. Top row (left to right), Nonenhanced CT scan, nonenhanced T1-weighted MRI, T2-weighted MRI, and fluid-attenuated inversion recovery (FLAIR) MRI. Second row from top (left to right): Diffusion-weighted MRI, apparent diffusion coefficient (ADC) map, and axial and coronal contrast-enhanced T1-weighted MRIs. Third row from top: Perfusion-weighted MRIs show increased flow in the caudate and putamen but not in the other areas of abnormal signal intensity. Bottom row: Axial spectroscopic image shows the 2 regions of interest in the right caudate corresponding to the multivoxel spectra. Note the large, infiltrating mass centered in the right basal ganglia and extending to the right frontal lobe, temporal lobe, and insula. Image shows thick peripheral enhancement and central necrosis. Multivoxel spectroscopy demonstrates decreased N-acetylaspartate (NAA), elevated choline, and elevated lactate values in thecaudate. Magnetic Resonance Imaging
MRI has increased the sensitivity and specificity in imaging astrocytomas. With the advent of the new techniques (MRS, perfusion-weighted imaging, and diffusion-tensor imaging), specificity has further improved.[6, 7, 8]
Magnetic resonance spectroscopy
MRS allows cerebral metabolites to be assessed by suppressing the signal of water and by interrogating for entities, including NAA, Cho, creatine (Cr), lactate, and lipids. The 2 main MRS techniques are single-voxel spectroscopy (shown in the image below) and chemical-shift imaging.[9, 10] Single-voxel spectroscopy is used to detect the signal from a single region during 1 measurement. Chemical-shift imaging uses additional phase-encoding pulses to obtain signals.
Normal pattern on single-voxel magnetic resonance spectroscopy (MRS) of 4 key molecules show relative heights and typical values. N-acetylaspartate (NAA) peaks at 2.0 ppm (tallest peak). Creatine (Cr) peaks at 3.0 ppm. Choline (Cho) peaks at 3.2 ppm. Myo-inositol (MI) peaks at 3.6 ppm.
Image show no evidence of recurrence in a 37-year-old woman with a history of grade 3 astrocytoma who underwent resection 7 years ago. Coronal contrast-enhanced T1-weighted, diffusion-weighted, and apparent diffusion coefficient (ADC) imaging were initially performed, followed by multivoxel magnetic resonance spectroscopy (MRS) and perfusion imaging. Because findings on cross-sectional imaging were not conclusive, additional studies were performed. MRS shows low levels of N-acetylaspartate (NAA), creatine (Cr), and choline (Cho) in an area of encephalomalacia, with a normal spectrum in the rest of the brain. Perfusion imaging shows no increased activity in the area of concern. These findings are compatible with gliosis. With cerebral gliomas, MRS is used to assess the spectral pattern, metabolite intensities, and ratios to help grade the tumor and/or predict treatment response (see the image below). MRS can also help in evaluating for tumoral recurrence and treatment response. The intensity of NAA is correlated with neuronal density and viability. Cho is involved in the turnover of cell membranes and neurotransmitters. Cr serves as a reserve for high-energy phosphates in the cytosol of neurons. Cerebral lactate is always abnormal and indicates ineffective cellular oxidative metabolism. Free lipids are present in areas of necrosis. Compared with normal tissues, cerebral gliomas consistently show lowered NAA intensity, elevated Cho (indicating increased membrane metabolism), and a stable or reduced Cr concentration. An Ins peak is described in certain low-grade tumors.
Perfusion-weighted imaging
Perfusion-weighted imaging involves several image acquisitions during the first pass of a bolus of contrast agent. This method allows the imager to determine the relative cerebral blood volume (rCBV). In general, the greater the rCBV, the higher the grade of tumor. Lack of notable flow indicates a nonneoplastic etiology with abnormal signal intensity, such as demyelination. Of note, mixed oligodendrogliomas can have low rCBV. Besides the prognostic information it provides, perfusion-weighted imaging can increase the yield of brain biopsy and help in differentiating recurrent neoplasm from radiation necrosis.
Diffusion-tensor imaging
Diffusion-tensor imaging is an experimental sequence that allows the imager to evaluate the structure and orientation of the white matter tracts. This sequence takes advantage of the fact that myelin restricts diffusion of water molecules in directions perpendicular to the fiber orientation. This sequence can help in determining whether neoplasm involves white matter pathways, improving the precision of surgical planning and the placement of radiation ports.
Functional MRI
Although not used in the diagnosis of astrocytomas, functional MRI (fMRI) deserves mention because it can be an important part of presurgical planning. A blood oxygen–dependent sequence is applied as the patient performs various tasks involving motor, sensory, visual, auditory, and language functions. Increased blood flow to a part of the brain is correlated with increased metabolic activity. The results are used to determine whether tumor involves vital structures (eloquent areas), a finding that may possibly affect surgical decisions.
Astrocyte characteristics on MRI
Low-grade astrocytomas are typically hyperintense on T2-weighted images. On T1-weighted images, most low-grade astrocytomas are hypointense relative to white matter. Contrast enhancement may be absent or, at best, mild. Exceptions include the mural nodule of pilocytic astrocytoma and the strong heterogeneous enhancement of pleomorphic xanthoastrocytomas. Astrocytomas are often associated with enhancement of the adjacent dura and meninges, giving the dural-tail appearance. MRS may show an elevated Cho peak and decreased NAA peak. An elevated Cho-Cr ratio or a depressed NAA-Cr ratio suggests tumor. This holds true for all high-grade tumors and many, but not all, low-grade tumors. (Some low-grade tumors may not have an elevated Cho peak.) Perfusion MRI studies fail to demonstrate increased rCBV.
Grade III astrocytomas often invade structures without destroying them, causing their ill-defined borders. The mass is inhomogeneous and bright on T2-weighted images. Surrounding edema and/or tumor infiltration is usually appreciated. Enhancement is usually seen. Perfusion MRI demonstrates increased relative cerebral flow volume.
Grade IV astrocytomas (GBM) are usually discovered as bulky disease, and necrosis is a hallmark of this grade. These lesions usually enhance peripherally, in a nodular and irregular manner, and they cause a large amount of mass effect and edema. These tumors often cross the corpus callosum, giving them a typical butterfly shape. Areas of hemorrhage and necrosis are common, and spectroscopy demonstrates high Cho, high lactate, high lipid, and low NAA values. Short–echo time (TE) studies demonstrate an absent or low myo-inositol peak. Perfusion studies demonstrate elevated rCBV.
See MRI images of brain astrocytomas below.
Pilocytic astrocytoma in a 20-year-old man. Top row (left to right), sagittal, coronal, and axial contrast-enhanced T1-weighted MRIs. Bottom row: Axial fluid-attenuated inversion recovery (FLAIR), diffusion, and apparent diffusion coefficient (ADC) images. Note the cystic mass with an intensely enhancing mural nodule in the inferior cerebellar vermis, as well as the mass effect on the brainstem, upper cervical cord, cerebellum, and fourth ventricle.
Grade II astrocytoma in a 30-year-old man. Nonenhanced T2-weighted MRI shows a well-circumscribed area of increased signal intensity in the left temporal lobe.
Grade II astrocytoma. Left, Fluid-attenuated inversion recovery (FLAIR) image demonstrates an area of increased signal intensity in the parietooccipital region. Right, Perfusion MRI demonstrates decreased relative cerebral blood volume (rCBV), consistent with a low-grade neoplasm. The final pathologic diagnosis was a grade II astrocytoma.
Grade III astrocytoma in a 33-year-old woman. Top row (left to right), Axial nonenhanced and contrast-enhanced T1-weighted, proton density–weighted, and fluid-attenuated inversion recovery (FLAIR) MRIs. Bottom row (left to right), Sagittal nonenhanced and contrast-enhanced T1-weighted MRIs, axial diffusion-weighted images, and axial apparent diffusion coefficient (ADC) map. T1-weighted images demonstrate a well-defined area of mixed signal intensity in the right parietal lobe extending to the corpus callosum with adjacent vasogenic edema. Mixed areas represent hemorrhage. Contrast-enhanced images show minimal enhancement of the lesion. Also appreciated is dural enhancement secondary to previous intervention.
Grade III astrocytoma in a 71-year-old man. Top row (left to right), Axial nonenhanced and contrast-enhanced T1-weighted, proton density–weighted, and fluid-attenuated inversion recovery (FLAIR) MRIs. Bottom row (left to right), Sagittal nonenhanced and contrast-enhanced T1-weighted MRIs, axial diffusion-weighted images, and axial apparent diffusion coefficient (ADC) map. Images show a cystic, well-defined lesion in the left parietal region with surrounding vasogenic edema and a thick rim enhancement on enhanced images. Diffusion and ADC images shows no evidence of acute restriction. Recent studies
In a study, Bing et al concluded that first-pass perfusion MRI is a quick and useful way to differentiate between pilocytic astrocytomas and hemangioblastomas. They also maintained that the maximum rCBV (rCBVmax), defined as the ratio between the CBVmax in tumor tissue and the CBV in healthy, contralateral white matter, is indicative of the tumor type. Because pilocytic astrocytomas and hemangioblastomas can present the same morphologic characteristics on conventional MRI sequences, the authors studied 11 patients with pilocytic astrocytomas and 8 with hemangioblastomas, who underwent first-pass perfusion MRI to differentiate the tumors.[11]
The investigators found that the difference between the rCBVmax of pilocytic astrocytomas (rCBVmax = 1.19 ± 0.71, range 0.6-3.27) and that of hemangioblastomas (rCBVmax = 9.37 ± 2.37, range 5.38-13) was significant. The first-pass curve crossed the baseline, corresponding to vascular permeability problems in pilocytic astrocytomas and hemangioblastomas.
Nuclear Imaging
Primary brain tumors generally have increased glucose metabolism on [18F]fluorodeoxyglucose (FDG) PET scan studies (see the image below).[12, 13] The degree of metabolic activity is correlated with the grade of the tumor and the patient's prognosis. Low-grade tumors may demonstrate little to no increased uptake, whereas grade IV lesions often have uptake that overshadows that of the gray matter.
A, Image in a patient after resection of a left frontoparietal, high-grade astrocytoma. Positron emission tomography (PET) demonstrates increased activity in this region, consistent with recurrence. B, Image in another patient being evaluated for recurrence of a high-grade astrocytoma. Image shows no abnormally increased activity to suggest recurrence. The heterogeneous nature of grade III or IV lesions is a specific feature for which PET scanning is helpful. Obtaining samples from active areas is of vital importance for accurate grading. PET scanning can be used to guide stereotactic biopsy to obtain a representative sample.
PET scanning has also been used to assess the response to therapy, and it can be used to detect transformation of a low-grade lesion to a high-grade one.[1]
Angiography
Angiography is being used in several ongoing trials to assess the intratumoral treatment of grade III and IV astrocytoma.
Khan MK, Hunter GK, Vogelbaum M, Suh JH, Chao ST. Evidence-based adjuvant therapy for gliomas: Current concepts and newer developments. Indian J Cancer. Apr-Jun 2009;46(2):96-107. [Medline].
Riemenschneider MJ, Reifenberger G. Astrocytic tumors. Recent Results Cancer Res. 2009;171:3-24. [Medline].
Linscott LL, Osborn AG, Blaser S, Castillo M, Hewlett RH, Wieselthaler N, et al. Pilomyxoid astrocytoma: expanding the imaging spectrum. AJNR Am J Neuroradiol. Nov 2008;29(10):1861-6. [Medline].
Weissleder R. Primer of Diagnostic Imaging. 3rd ed. St Louis, Mo: Mosby-Year Book;2003.
Grainger R, Allison D, Andreas A, et al, eds. Grainger & Allison's Diagnostic Radiology: A Textbook of Medical Imaging. 4th ed. London: Churchill Livingstone;2001.
Harrer JU, Parker GJ, Haroon HA, et al. Comparative study of methods for determining vascular permeability and blood volume in human gliomas. J Magn Reson Imaging. Nov 2004;20(5):748-57.
Hsu YY. Metabolism, perfusion and diffusion MR images of cerebral gliomas. Acta Neurol Taiwan. 2002;11:173-180.
Runge V. Clinical MRI. Philadelphia, Pa: WB Saunders;2002.
Meyerand ME, Pipas JM, Mamourian A, et al. Classification of biopsy-confirmed brain tumors using single-voxel MR spectroscopy. AJNR Am J Neuroradiol. Jan 1999;20(1):117-23.
Tien RD, Lai PH, Smith JS, Lazeyras F. Single-voxel proton brain spectroscopy exam (PROBE/SV) in patients with primary brain tumors. AJR Am J Roentgenol. Jul 1996;167(1):201-9.
Bing F, Kremer S, Lamalle L, Chabardes S, Ashraf A, Pasquier B, et al. [Value of perfusion MRI in the study of pilocytic astrocytoma and hemangioblastoma: preliminary findings]. J Neuroradiol. May 2009;36(2):82-7. [Medline].
Alavi JB, Alavi A, Chawluk J, et al. Positron emission tomography in patients with glioma. A predictor of prognosis. Cancer. Sep 15 1988;62(6):1074-8.
Hanson MW, Glantz MJ, Hoffman JM, et al. FDG-PET in the selection of brain lesions for biopsy. J Comput Assist Tomogr. Sep-Oct 1991;15(5):796-801.

