GBM versus solitary metastasis

It is clinically important to differentiate GBM from a solitary brain metastasis because medical staging, surgical planning, and therapeutic decisions are vastly different for each tumor type.

On conventional MR images, HGGs and solitary metastatic brain tumors often display similar signal-intensity characteristics and contrast-enhancement patterns. Advanced MR imaging techniques may help in differentiating solitary brain metastasis from high-grade glial tumors such as GBM.

Peritumoral ADC

GBMs and metastatic tumors often contain heterogeneous signal intensity secondary to necrosis and susceptibility artifacts. As a result of this heterogeneity, DW imaging metrics obtained from the tumor can be imprecise or inaccurate. Lee and colleagues demonstrated that tumoral ADC showed no statistically significant difference between the two groups (GBM and metastasis). On the other hand, the mean minimum ADC value in the peritumoral region of GBMs (1.149 ± 0.119 [standard deviation]) was significantly lower than that in metastases (1.413 ± 0.147) ( P <.05). The mean peritumoral ADC ratio (minimum ADC in the peritumoral edema of the affected hemisphere divided by that from the normal white matter in the contralateral hemisphere) was also significantly lower in GBMs (1.466 ± 0.24) than in metastases (1.829 ± 0.25) ( P <.05). These results support the hypothesis that minimum ADC values can detect neoplastic cell infiltration in peritumoral edema in patients with GBM, distinguishing these tumors from solitary brain metastasis.

Differentiation of vasogenic edema from infiltrative edema has been attempted using DW imaging, based on the premise that water diffusivity is facilitated to a greater degree in vasogenic edema surrounding brain metastasis than in infiltrative edema surrounding GBMs, because of a lack of intervening tumor cells in the former. However, the authors believe that ADC measurements in the peritumoral areas of GBM and metastasis cannot be used to distinguish between these lesions, as a large overlap is often demonstrated. This finding is in agreement with previous studies, which have shown that peritumoral ADC is not useful for distinguishing between GBMs and metastatic tumors.

Peritumoral diffusion tensor imaging metrics

Areas of extensive edema on T2-weighted imaging surround both GBMs and metastatic tumors. In metastatic tumors, this peritumoral edema is assumed to be composed of pure water, whereas for GBMs the edema is assumed to contain tumor cells that have infiltrated into the tissue.

Diffusion tensor (DT) imaging is an advanced MR technique that describes the movement of water molecules by using 2 metrics, mean diffusivity (MD) and FA, which represent the magnitude and directionality of water diffusion, respectively. DT imaging may be able to detect changes in the white matter surrounding malignant gliomas that are not visible on cMR imaging. In a study by Lu and colleagues the peritumoral MD of metastatic lesions was significantly greater (average 0.798 ± 0.109) than that of gliomas (average 0.622 ± 0.111) ( P <.005). The average FA was 0.181 ± 0.049 around metastatic lesions, whereas the average FA around HGGs was 0.248 ± 0.063.

Although peritumoral MD could be used to distinguish HGGs from metastatic tumors, peritumoral FA demonstrated no statistically significant difference.

¹ H-MRS

To differentiate a solitary brain metastasis from GBM, the voxel should be placed in the peritumoral area. No evidence of tumor infiltration is demonstrated in the peritumoral vasogenic edema surrounding a metastatic brain lesion. No elevation of the Cho, Cho/creatine (Cho/Cr), and Cho/NAA ratios will be demonstrated in the spectrum ( Fig. 1 A, B ). On the other hand, evidence of tumor infiltration outside the tumor nodule may be demonstrated in the peritumoral area surrounding a GBM (see Fig. 1 C, D).

Magnetic resonance spectroscopy (MRS) of perilesional edema. ( A , B ) A 66-year-old woman diagnosed with breast cancer metastasis. There is a solitary nodular metastatic lesion in the left parietal lobe, surrounded by vasogenic edema ( A : axial T2). Spectrum from the perilesional edema ( B ) demonstrates no signs of tumor infiltration: there is no elevation of the choline (Cho) peak, Cho/creatine (Cr) ratio, and Cho/ N -acetylaspartate (NAA) ratio. ( C , D ) A 39-year-old woman presenting with aphasia. There is a left parietal nodular lesion, isointense on T2 ( C ) surrounded by vasogenic edema. Spectrum from the perilesional “edema” ( D ) demonstrates significant elevation of the Cho peak, Cho/Cr ratio, and Cho/NAA ratio, consistent with tumor infiltration. Final diagnosis: glioblastoma multiforme (GBM).

Beware that high Cho peak, Cho/Cr ratio, and Cho/NAA ratio in the peritumoral region surrounding a GBM may not be demonstrated in the spectrum, as these changes are related to the number of tumor cells that have infiltrated the peritumoral area ( Fig. 2 ). If no elevation of the Cho peak is seen in the peritumoral area, a high-grade primary tumor cannot be ruled out.

MRS of the perilesional edema surrounding a GBM. A 50-year-old man presenting with headaches and paresthesia on the right. There is a left frontoparietal GBM presenting with irregular enhancement ( A : axial T1 with contrast). Spectrum from the perilesional area ( B ) shows no signs of tumor infiltration. The Cho peak is smaller than the Cr peak.

Dynamic susceptibility contrast MR imaging

Peritumoral CBV

When a focal brain lesion presents with high blood volume, GBM and metastasis should be both considered in the differential diagnosis. In metastatic brain lesions, the area with high blood volume will correspond exactly to the tumoral-enhancing nodule. On the other hand, GBMs tend to have angiogenesis beyond the contrast-enhancing portion of the tumor, so that high blood volume may be demonstrated outside the tumor nodule ( Fig. 3 ).

Peritumoral relative cerebral blood volume (rCBV). A 67-year-old woman presenting with headaches. There is a nodular enhancing lesion close to the posterior margin of the left ventricle ( A : axial T1 with contrast) that was diagnosed as a GBM. High blood volume ( B ) can be seen extending beyond the limits of the enhancing lesion in A , related to vascular proliferation as well as tumor infiltration into the peritumoral tissue.

In a study of 75 brain tumors, peritumoral CBV values (rCBVp) were more effective than tumoral rCBV (rCBVt) in differentiation of metastasis from both LGG and HGG. rCBVp cutoff values of 1.1 and 1.2 were used for the differentiation.

Percentage of signal-intensity recovery (curve analysis)

Percentage recovery analysis of time-intensity curves from GE-EPI DSC PW imaging has been proposed to distinguish GBMs from metastasis by assessing the profound difference in capillary permeability between the 2 tumor types with increased capillary permeability in brain metastasis.

The percentage of signal-intensity recovery (PSR) is a hemodynamic variable derived from DSC perfusion MR imaging that provides additional information on tumor vasculature. PSR is defined as the percentage of signal-intensity recovery relative to the precontrast baseline at the end of the first pass, and is influenced by the amount of contrast agent leaked during the first pass, thus reflecting alteration in capillary permeability.

Cha and colleagues demonstrated that both the minimum and average percentages of signal-intensity recovery were significantly greater for patients with GBM compared with the values for patients with metastasis ( P <.05). When the average PSR in the contrast-enhancing lesion was more than 82% and less than 66%, the prediction of GBM and single brain metastasis, respectively, had specificity of 100% ( Fig. 4 ).

Percentage of signal intensity recovery: mets × GBM. A low percentage of signal intensity recovery (PSR) of the time-intensity curve is demonstrated in the perfusion study in metastasis ( A ). In the GBM a greater PSR is usually seen ( B ).

Beware: Some GBMs with significant disruption of the blood-brain barrier (BBB) will demonstrate a very low PSR of the time-intensity curve, resembling metastasis ( Fig. 5 ).

Percentage of signal intensity recovery: drawback. GBM in the left frontoparietal region presenting with very high blood volume ( A : CBV map). The time-intensity curve ( B ) demonstrates a very low PSR, resembling metastasis (compare with Fig. 4 A).

Dynamic contrast-enhanced MR imaging

Permeability is typically high in aggressive tumors such as metastasis and GBM, owing to significant disruption of the BBB. There is greater permeability in metastasis than in GBM because metastases have no BBB, whereas GBMs have a “damaged” BBB.

GBM versus lymphoma

Despite some characteristic conventional MR imaging findings, it may be difficult or even impossible to distinguish cerebral lymphomas from GBM.

Accurate preoperative differentiation between these 2 tumors is important for appropriate treatment.

GBMs are usually treated with surgical resection plus radiation therapy and chemotherapy, whereas lymphoma is not treated with surgery.

ADC and FA

Lymphomas demonstrate low ADC values ( Fig. 6 A–C ), a finding consistent with restricted water diffusion related to high tumor-cell density. By contrast, HGGs are relatively hyperintense to gray matter on both trace DW imaging and ADC maps ( Fig. 7 B ), findings consistent with elevated diffusivity. Prior studies have shown statistically significant differences in ADC between the cerebral lymphoma and GBM. However, GBMs with restricted water diffusion have also been reported. Therefore, discrimination of lymphoma from some GBMs may be difficult.

GBM versus lymphoma. A 37-year-old man, human immunodeficiency virus (HIV)-positive, diagnosed with lymphoma involving the basal ganglia and periventricular regions ( A : axial T1 with contrast). The solid areas present with low apparent diffusion coefficient (ADC) values ( B : ADC map). The ADC ratio (ADC from the lesion divided by the ADC from the normal white matter) is less than 1 ( C ). Spectrum from the solid thalamic lesion on the left ( D , E ) demonstrates high Cho peak along with high lipid-lactate peaks. There is no elevation of the blood volume ( F : CBV map) and only minimal elevation of the permeability ( G : permeability map), which favors lymphoma over GBM. The minimal elevation of permeability in this lymphoma could represent a limitation of trying to estimate permeability without pharmacokinetic modeling, and further confirmation is necessary.

GBM versus lymphoma. There is a heterogeneously enhancing GBM ( A : axial T1 with contrast) involving the parietal lobes and the cingulate gyrus bilaterally, presenting with high ADC ( B : ADC map). A prominent lipid peak is demonstrated in the spectrum from the necrotic area ( C ). Very high blood volume ( D : CBV map) and significantly high permeability ( E : permeability map) are demonstrated, very consistent with the diagnosis of GBM.

A study by Toh and colleagues demonstrated that the FA and ADC of lymphomas were significantly lower than those of GBMs. Cutoff values to differentiate lymphomas from GBM were 0.192 for FA, 0.33 for FA ratio, 0.818 for ADC, and 1.06 for ADC ratio. Accuracy of 100% was reached in the distinction between lymphoma and GBM using a cutoff value of 1.06 for the ADC ratio (see Fig. 6 C).

The specificity and accuracy of ADC were higher than those of FA in differentiating the two.

ADC measurements are also useful in the distinction between lymphoma and GBM infiltrating the corpus callosum.

¹ H-MRS

When lipids are demonstrated in a solid-appearing tumor, lymphoma should be considered (see Fig. 6 D, E).

Lipids are also typically seen in the spectrum of the GBM (see Fig. 7 C), related to the presence of necrosis.

DSC and DCE MR imaging

CBV and cerebral blood flow are much lower in lymphomas than in GBMs.

Angiogenesis is not a prominent finding in lymphomas, so perfusion may not be elevated (see Fig. 6 F).

In the authors’ experience, despite being an aggressive tumor, lymphomas may present with minimal or mild elevation of the permeability (see Fig. 6 G). On the other hand, high blood volumes and significantly high permeability are typically demonstrated in high-grade glial tumors such as GBM (see Fig. 7 D, E).

Gliomatosis cerebri versus lymphomatosis cerebri

Intracranial gliomas may present as infiltrative lesions known as gliomatosis cerebri. These tumors usually infiltrate at least 2 or 3 lobes, sometimes in a bilateral and fairly symmetric fashion, usually with no or minimum mass effect.

Lymphomas may also present as infiltrative tumors known as lymphomatosis cerebri. This rare presentation of intracranial primary lymphomas may resemble gliomatosis cerebri on cMR imaging. Discrete or no enhancement is demonstrated in both tumors.

ADC measurements are not helpful in the differentiation of these lesions. Restricted diffusion, typically seen in lymphomas, is mild or absent in lymphomatosis cerebri.

¹ H-MRS may be helpful, and a high mI peak has been described as the most characteristic spectral finding in gliomatosis grade II, especially if the Cho peak is not elevated ( Fig. 8 A, B ). On the other hand, lymphomatosis cerebri is characterized by reduction of the NAA and mI peaks, along with elevation of the Cho and lipid peaks (see Fig. 8 C, D).

Lymphomatosis versus gliomatosis cerebri. Gliomatosis cerebri infiltrating the white matter bilaterally ( A : coronal T2). The spectrum ( B ) demonstrates reduction of the NAA and Cho peaks along with elevation of the myo -inositol (mI) peak. Lymphomatosis cerebri infiltrating the frontal and parietal lobes bilaterally, mostly on the left ( C : axial fluid-attenuated inversion recovery [FLAIR]). A prominent lipid peak, along with elevation of the Cho peak and reduction of the NAA and mI peaks, are demonstrated in the spectrum ( D ).

([ C , D ] Courtesy of Leonardo Avanza MD, Vitória ES, Brazil.)

Medulloblastoma versus ependymoma

Medulloblastoma, a highly malignant neoplasm, is the most common neoplasm of the posterior fossa in children. The tumor usually arises at the midline within the vermis, and exhibits growth into the fourth ventricle.

Medulloblastomas and ependymomas may look very similar on cMR imaging. Accurate diagnosis is crucial, as treatment strategies and prognosis are different.

ADC

ADC values are significantly lower in medulloblastomas than in all other posterior fossa tumors ( P <.001) related to high cell density ( Fig. 9 A, B ).

Medulloblastoma versus ependymoma: ADC map. ( A , B ) Medulloblastoma. ( C , D ) Ependymoma. The solid portions of medulloblastomas and ependymomas may be isointense on the T2 sequence ( A , C ). Cysts may be seen in both tumors ( arrows in A , C ). Mass in the cerebellopontine angle is very characteristic of ependymomas ( C ). Restricted diffusion is classically seen in the medulloblastoma ( B : ADC map) because of high cellularity. There is no restricted diffusion in the classic ependymoma ( D : ADC map).

Diffusion is not typically restricted in the classic ependymomas (see Fig. 9 C, D), but may be restricted in the anaplastic ones, making distinction between these tumors and medulloblastomas more difficult.

¹ H-MRS

Medulloblastomas usually demonstrate a significant elevation of the Cho peak, Cho/Cr ratio, and Cho/NAA ratio in the spectrum, related to the high cell density and reflecting its malignant nature ( Fig. 10 A ). Majós and colleagues demonstrated that significantly elevated levels in the Cho peak could be used to distinguish primitive neuroectodermal tumors (PNET) from non-PNET tumors with a high degree of accuracy (94%).

Medulloblastoma versus ependymoma: ¹ H-MRS. ( A ) Spectrum from a medulloblastoma demonstrates significant elevation of the Cho peak, along with presence of lipids, lactate, and taurine (tau) peaks. ( B ) Spectrum from the posterior fossa ependymoma shown in Fig. 9 C demonstrates a very high mI peak.

Also, spectrum with a short echo time (TE) show a significantly elevated taurine concentration at 3.3 ppm in patients with medulloblastoma when compared with other tumors (see Fig. 10 A).

Ependymomas may show very high mI in spectrum obtained with short TE (30 milliseconds), allowing distinction between these tumors and medulloblastomas (see Fig. 10 B).

Dynamic susceptibility contrast MR imaging

In the authors’ experience, some medulloblastomas present with high blood volumes and others with low blood volumes.

Ependymomas generally demonstrate markedly elevated rCBV ( Fig. 11 A–C ) and, unlike many other glial neoplasms, a poor return to baseline that may be attributable to fenestrated blood vessels and an incomplete BBB (see Fig. 11 C).

Ependymoma versus medulloblastoma: dynamic susceptibility contrast (DSC) study. There is a heterogeneously enhancing ependymoma arising from the fourth ventricle extending to the left cerebellopontine angle (CPA) ( A : axial T1 with contrast). The lesion has very high blood volume ( B : rCBV map). A very low PSR of the time-intensity curve is demonstrated ( C, arrows ) related to fenestrated blood vessels and an incomplete blood-brain barrier. See also Refs.

Dynamic contrast-enhanced MR imaging

In the authors’ experience, some medulloblastomas have demonstrated no significant elevation of the permeability ( Fig. 12 A–C ).

Medulloblastoma: dynamic contrast-enhanced (DCE) study. Adult patient diagnosed with medulloblastoma. There is a solid enhancing lesion compromising the cerebellar vermis ( A ) presenting with no significant elevation of the permeability ( B : permeability map; C : signal intensity–phase curves). Mild elevation of the permeability in this case could represent a limitation of trying to estimate permeability without pharmacokinetic modeling, and further confirmation is necessary.

Permeability in ependymomas is highly variable, with some ependymomas presenting with high permeability ( Fig. 13 A, B ) and some with no significant elevation of the permeability (see Fig. 13 C, D) (authors’ personal communication).

Ependymoma: DCE study. ( A , B ) Ependymoma in an adult patient. There is a nodular enhancing lesion arising from the fourth ventricle, extending to the left CPA ( A : axial T1 with contrast), presenting with high permeability ( B : permeability map). ( C , D ) A 3-year-old boy diagnosed with ependymoma. There is an enhancing lesion ( A : axial T1 with contrast) extending from the fourth ventricle to the CPA on the left, presenting with much lower permeability ( D : permeability map) compared with the patient in A , B .

Hemangioblastoma versus pilocytic astrocytoma

Hemangioblastomas (HGBLs) are the most common primary brain tumors in the posterior fossa of adult patients. The lesion arises in the cerebellar parenchyma, and usually presents with cystic and solid components ( Fig. 14 ). HGBL may resemble a pilocytic astrocytoma (PA) on cMR imaging.

Hemangioblastoma (HGBL). A 57-year-old woman presenting with vertigo, ataxia, and headaches in the last 2 months. There is a cystic-solid lesion in the cerebellum. Flow voids consistent with prominent vessels are demonstrated within the solid component of the lesion, suggesting the diagnosis of HGBL ( A : axial T2). There is no restricted diffusion within the cysts as well as in the solid portion of the tumor ( B : ADC map). There is no significant elevation of the Cho peak in the spectrum ( C ), as this is a low-grade tumor. Lipids are demonstrated, related to the presence of cysts. ( D ) Spectrum from a pilocytic astrocytoma for comparison demonstrates a prominent Cho peak. There is apparent high blood volume ( E : rCBV map) and significantly high permeability ( F : permeability map; G : signal intensity–phase curves region of interest [ROI] 1, venous sinus and ROI 2, solid lesion), which may both be related to the presence of macrovessels.

Flow voids consistent with prominent blood vessels may be demonstrated in the solid portion of HGBLs (see Fig. 14 A), suggesting the diagnosis.

ADC

ADC measurements are related to tumor-cell density and tumor grade. Many studies have demonstrated an inverse relationship between ADC values and tumor grade, with lower ADCs found in higher-grade tumors ( Figs. 15 and 16 ) in comparison with lower-grade tumors ( Fig. 17 ).

Grade III glioma. A 41-year-old woman presenting with depression, headaches, and panic attacks. There is a solid mass in the left temporal-frontal region ( A : axial T2) with a faint nodular enhancement in its anterior portion ( B : axial T1 with contrast). There is an area of restricted diffusion within the lesion ( circle in C : ADC map) that corresponds to the area of lowest signal intensity on the T2 image ( A ), indicating high cell density. A prominent Cho peak is demonstrated in the spectrum ( D ). There is elevation of the blood volume ( E : CBV map). No significant elevation of the permeability is demonstrated ( F : permeability map; G : signal intensity–phase curves). Pathology was consistent with grade III (anaplastic) oligoastrocytoma.

Grade IV glioma. A 64-year-old woman found unconscious. There is a large heterogeneous lesion involving the left frontal and parietal lobes ( A : axial T2). An area with restricted diffusion ( B : ADC map) is demonstrated within the lesion, compatible with high cell density. Spectrum ( C ) shows a prominent lipid peak, due to extensive necrosis. No elevation of the Cho is demonstrated. The lesion presents with heterogeneous enhancement ( D : axial T1 with contrast). Very high blood volume is demonstrated ( E : CBV map) extending beyond the limits of the enhancing tumor ( D ), characterizing tumor infiltration beyond the enhancing lesion. Permeability is also very high ( F , G : permeability maps).

Low-grade (grade II) astrocytoma. A 5-year-old boy who presented with 1 seizure episode at the age of 7 months. The study was done because of motor delay. A solid lesion is demonstrated in the right frontal lobe, presenting with no significant enhancement ( A : axial with contrast). There is increased ADC in the mass ( B : ADC map) suggesting a low-grade tumor. Spectrum ( C ) demonstrates a prominent mI peak, very typical for grade II astrocytomas. Perfusion is low ( D ) and there is no elevation of the permeability ( E ). The findings from the conventional and functional studies were consistent with low-grade tumor (confirmed as grade II astrocytoma).

Recent studies have shown that minimum ADCs may facilitate accurate grading of astrocytic tumors, because regions exhibiting the minimum ADC correspond to the highest-grade glioma foci within heterogeneous tumors.

Beware: restricted diffusion may be seen in PA, a grade I tumor ( Fig. 18 ).

Pilocytic astrocytoma (PA). A 9-year-old boy presenting with epilepsy and right-side hemiparesis. The solid portion of the PA is isointense on T2 ( A ) and presents with very high signal intensity on diffusion-weighted (DW) imaging ( B ) characterizing restricted diffusion. The spectrum ( C ) demonstrates a prominent Cho peak along with elevation of lipids and lactate (Lip-lact), and reduction of the NAA and Cr peaks.

¹ H-MRS

¹ H-MRS may indicate tumor grade with more accuracy when compared with a blind biopsy, because it assesses a larger amount of tissue than is usually excised at biopsy.

Tumors are commonly heterogeneous, and their spectrum may vary depending on the region sampled by ¹ H-MRS. Hence, the ROI chosen for analysis will have a large influence on the results, and multivoxel (MV) spectroscopy is generally considered preferable because it allows metabolic heterogeneity to be evaluated. ¹ H-MRS is considered 96% accurate in differentiating LGGs from HGGs.

There is a high correlation between in vivo concentration of Cho in brain tumors and in vitro tumor proliferation markers. Compared with low-grade gliomas (LGGs) (see Fig. 17 C, D), statistically significantly higher Cho/Cr, Cho/NAA, and rCBV have been reported in HGGs (see Fig. 15 D, E). although threshold values of metabolite ratios for grading of gliomas are not well established. Cho/Cr is the most frequently used ratio for grading gliomas.

Although Cho is related to tumor-cell density and tumor grade, grade IV GBM usually presents with lower levels of Cho (see Fig. 16 C) than gliomas of grades III (see Fig. 15 D) and II (see Fig. 17 C). This discrepancy may be due to the presence of necrosis in GBMs, because necrosis is associated with a prominent lipid peak along with reduction of all other metabolites. Lipids may be the only spectral finding in a GBM.

Useful information on tumor grade may be acquired by using a short-TE (30–35 milliseconds) ¹ H-MRS to assess mI. In low-grade tumors, the mI/Cr ratio is typically higher (see Fig. 17 C) than in high-grade tumors (see Fig. 15 D). According to Howe and colleagues, high mI is characteristic of grade II astrocytomas. Increased mI levels have been reported to be useful in identifying low-grade astrocytomas in which the Cho/Cr ratio is not altered.

DSC MR imaging

The rCBV, the most robust and standard hemodynamic variable derived from DSC perfusion MR imaging, has been shown to correlate with grading of pure astrocytomas.

CBV is a marker of angiogenesis, tumor grade, and survival. CBV is related not only to vascular density but also to tumor-cell density.

Law and colleagues have demonstrated that rCBV is better related to tumor biology than Cho/Cr and Cho/NAA ratios.

Tumor rCBV obtained with GE-EPI is significantly higher than that obtained with the SE-EPI technique in the HGGs. This is probably related to the higher sensitivity of the GE-EPI technique to large vessels, which can be clearly observed in the HGGs. The findings suggest that the GE-EPI technique is more useful for detecting low-versus high-grade gliomas than is the SE-EPI technique. However, SE-EPI may have a better performance than the GE-EPI technique in areas of greater susceptibility, such as close to the bone and surgical cavities.

Permeability can confound rCBV measurements, and correction for T1 and T2/T2* effects should be done before interpreting the results from the perfusion study.

Many different techniques have been advocated, although no standardized methods exist. This ambiguity can result in a wide range of values in the literature, and makes comparison of studies problematic.

Perfusion studies can also be done without gadolinium-based contrast agent (GBCA) injection using the arterial spin-labeling (ASL) technique. The results from this technique are related to the percentage of blood vessels and tumor cellularity, and may be useful in distinguishing between high-grade and LGGs. ASL currently can provide measurements of cerebral blood flow, and absolute quantification is more easily performed in ASL than in DSC MR imaging.

Using the SE-EPI technique, Lev and colleagues found that if the tumoral rCBV was more than 1.5 times the rCBV in the contralateral white matter, a HGG could be suggested with 100% sensitivity and 69% specificity. According to this study, none of the HGGs presented with low CBV. On the other hand, using the GE-EPI sequence, Law and colleagues demonstrated that if the measured rCBV in the tumor lesion was more than 1.75 times the rCBV in the contralateral white matter, a HGG could be suggested with 95% sensitivity and 57.5% specificity.

HGGs are characterized by high rCBV (see Figs. 15 E and 16 E; Fig. 21 A, B). However, a high proportion of LGGs can demonstrate foci of high rCBV ( Fig. 22 ).

Grade III glioma. A 28-year-old man presenting with 6-month history of generalized seizures. There is a solid lesion in the right occipital lobe, with heterogeneous enhancement ( A : axial T1 with contrast). The lesion presents with very high CBV ( B ) and very high permeability ( C : permeability map; D : signal intensity–phase curves). Tumor permeability (ROI 2) is almost 3 times that in the gray matter (ROI 4).

High rCBV in a grade II glioma. A 60-year-old woman presenting with seizures. There is a solid ill-defined lesion in the left frontal lobe that does not enhance ( A : axial T1 with contrast). High blood volume is demonstrated in the DSC MR study ( B : CBV map), a finding that can be seen in low-grade gliomas.

It is important to remember that both low-grade and high-grade oligodendrogliomas may present with high blood volume in the perfusion study, because of their unique “chicken-wire” vascularity. Low-grade oligodendrogliomas with high blood volumes tend to respond better to chemotherapy with procarbazine and vincristine.

Recently, susceptibility-weighted (SW) imaging has been described as a useful tool for the grading of noninvasive glioma. High rCBV values on perfusion imaging in tumors go hand in hand with evidence of blood products detected within the tumor matrix on SW imaging.

Beware: pilocytic astrocytomas may present with high blood volumes.

DCE MR imaging

Like PW MR imaging, permeability MR imaging, also known as DCE MR imaging, is performed through the use of GBCAs. However, it often uses more complex pharmacokinetic models (PKMs) than in PW imaging, and incorporates several assumptions that may potentially result in errors. It is insensitive to magnetic susceptibility and thus has a better performance than perfusion studies in the evaluation of lesions close to the skull base and close to the surgical cavity. The techniques utilized often result in conflicting demands on high temporal resolution, high signal to noise ratio, high spatial resolution, and wide anatomic coverage.

Permeability allows estimation of the BBB integrity through calculation of microvascular permeability. The volume-transfer constant between the capillary plasma space and the extravascular-extracellular space (EES), K ^trans , is the most widely used microvascular permeability metric. It is a potentially intractable combination of flow, permeability, and surface area. Despite the most rapid imaging techniques, it may not be possible to differentiate the contribution of each factor in the resultant K ^trans and, thus, its physiologic meaning can differ. Furthermore, variations in the DCE technique can also make a comparison of literature values of K ^trans difficult.

K ^trans of the brain parenchyma is insignificant, and consistent with noise.

K ^trans and CBV measure different vascular parameters, so areas of high CBV do not necessarily correspond to areas of high permeability. Strong relationships have been demonstrated between both rCBV and K ^trans and histologic grade in gliomas; however, the relationship appears to be stronger with rCBV.

More simple methods analysis that do not use PKM, such as qualitative or semiquantitative analysis of signal intensity–time curves, though physiologically less specific, can be beneficial. The GE Functool maximum (max) slope of increase is likely correlated with permeability, and subjective dynamic curve analysis can be beneficial.

ADC measurements

Some studies have shown that the minimum ADC values of tumors have preoperative prognostic importance in patients with malignant supratentorial astrocytoma.

ADC values are also predictive of clinical outcome in primary central nervous system lymphoma.

Lymphomas with more restricted diffusion have higher cell density and have a worse prognosis, so these tumors deserve more aggressive treatment.

¹ H-MRS

Majós and colleagues have demonstrated that a high-intensity value of the peaks at 0.98 and 1.25 ppm, attributed to lipids, correlates with tumoral necrosis and low survival in high-grade astrocytomas. These investigators also found that another region of the short TE spectrum, around 3.67 ppm, likely representing glucose, showed a direct correlation with patient survival. High metabolic activity and, consequently, poor prognosis correlate with depletion of glucose in the extracellular compartment and, accordingly, with low intensity of the resonances that represent this compound in the spectrum, centered at 3.67 ppm. The investigators concluded that ¹ H-MRS could be used to stratify prognostic groups in HGGs and that this prognostic assessment could be made by evaluating the intensity values of 2 points on the spectrum at short TE (0.98 and 3.67 ppm) and another 2 at long TE (0.98 and 1.25 ppm). Although short-TE ¹ H-MRS may be considered somewhat superior to long-TE ¹ H-MRS for prognostic assessment of HGGs, spectrum at both TEs may provide relevant information.

Oh and colleagues found a significantly shorter median survival time for patients with a large volume of metabolic abnormality, measured by ¹ H-MRS.

In grade IV gliomas, high relative volumes of regions with elevated Cho/NAA index were negatively associated with survival. Survival time was also negatively associated with high lactate and lipid levels and with low ADC within the enhancing volume.

DSC MR imaging

The predictive value of perfusion MR imaging for the prognosis of patients with glioma has been well documented. Lev and colleagues reported that in low-grade and HGGs, elevated rCBV was a stronger predictor of survival than the degree of enhancement.

According to Tzika and colleagues, the rCBV value was useful for distinguishing between progressive and stable tumors in pediatric patients with low-grade or high-grade glioma.

ADC

Low-grade tumors are typically characterized by relatively higher ADC values related to low cell density. During follow-up, foci of restricted diffusion related to high cellularity and malignant transformation may be demonstrated in these lesions ( Fig. 23 ). A calculated ADC ratio (tumoral ADC divided by the ADC in the contralateral white matter) of less than 1.0 is related to malignant transformation in LGGs.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Clinical Applications of Functional MR Imaging Breast Magnetic Resonance Imaging Clinical Applications of Diffusion Imaging in the Spine Magnetic Resonance Imaging of Rectal Cancer Multiparametric Magnetic Resonance Imaging of the Prostate MR Imaging of the Pelvic Floor

Stay updated, free articles. Join our Telegram channel

Join