ORIGINAL RESEARCH

䡲 NEURORADIOLOGY

Note: This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, use the Radiology Reprints form at the end of this article.

Low-Grade Gliomas: Do Changes in rCBV Measurements at Longitudinal Perfusion-weighted MR Imaging Predict Malignant Transformation?1 Nasuda Danchaivijitr, MD Adam D. Waldman, PhD Daniel J. Tozer, PhD Christopher E. Benton, BSc Gisele Brasil Caseiras, MD Paul S. Tofts, PhD Jeremy H. Rees, PhD H. Rolf Ja¨ger, MD

1 From the Institute of Neurology, University College London, Queen Square, London WC1 3BG, UK (N.D., A.D.W., D.J.T., C.E.B., G.B.C., P.S.T., J.H.R., H.R.J.); Department of Imaging, Charing Cross Hospital, London, UK (N.D., A.D.W.); Imperial College of Science Technology and Medicine, London, UK (A.D.W.); and Lysholm Department of Neuroradiology, National Hospital for Neurology and Neurosurgery, London, UK (G.B.C., H.R.J.). Received December 7, 2006; revision requested February 19, 2007; revision received July 31; accepted August 28; final version accepted September 28. Address correspondence to A.D.W. (e-mail: [email protected] ).

Purpose:

To prospectively perform longitudinal magnetic resonance (MR) perfusion imaging of conservatively treated lowgrade gliomas to determine whether relative cerebral blood volume (rCBV) changes precede malignant transformation as defined by conventional MR imaging and clinical criteria.

Materials and Methods:

All patients gave written informed consent for this institutional ethics committee–approved study. Thirteen patients (seven men, six women; age range, 29 – 69 years) with biopsy-proved low-grade glioma treated only with antiepileptic drugs were examined longitudinally with susceptibility-weighted perfusion, T2-weighted, fluid-attenuated inversion recovery, and high-dose contrast material– enhanced T1-weighted MR imaging at 6-month intervals to date or until malignant transformation was diagnosed. Student t tests were used to determine differences in rCBV values between “transformers” and “nontransformers” at defined time points throughout study follow-up.

Results:

Seven patients showed progression to high-grade tumors between 6 and 36 months (mean, 22.3 months), and disease in six patients remained stable over a period of 12–36 months (mean, 23 months). Transformers had a slightly (but not statistically significantly) higher group mean rCBV than nontransformers at the point of study entry (1.93 vs 1.31). In nontransformers, the rCBV remained relatively stable and increased to only 1.52 over a mean follow-up of 23 months. In contrast, transformers showed a continuous increase in rCBV up to the point of transformation, when contrast enhancement became apparent on T1-weighted images. The group mean rCBV was 5.36 at transformation but also showed a significant increase from the initial study at 12 months (3.14, P ⫽ .022) and at 6 months (3.65, P ⫽ .049) before transformation. Rates of rCBV change between two successive time points were also significantly higher in transformers than in nontransformers.

Conclusion:

In transforming low-grade glioma, susceptibility-weighted MR perfusion imaging can demonstrate significant increases in rCBV up to 12 months before contrast enhancement is apparent on T1-weighted MR images. 娀 RSNA, 2008

姝 RSNA, 2008

170

Radiology: Volume 247: Number 1—April 2008

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

A

dult supratentorial low-grade gliomas (World Health Organization [WHO] grade II, 2000) are a heterogeneous group of diffusely infiltrating primary brain tumors. They grow slowly for several years, but, at an unpredictable time, almost all progress to high-grade gliomas (WHO grade III or IV), which carry a poor prognosis. The management of low-grade gliomas remains controversial. Although in some centers, these lesions are treated aggressively at diagnosis, large population-based studies and prospective trials have not produced evidence of improved survival after radical surgery or early radiation therapy (1,2), particularly in young patients with well-controlled epilepsy at presentation. At our institution we have therefore adopted an expectant, “watch and wait” management policy of clinical and imaging surveillance at regular intervals. Further biopsy is performed at the first sign of clinical or imaging progression with a view to therapy, which may include surgical debulking, radiation therapy, or chemotherapy. A change in appearance at imaging frequently precedes clinical deteriora-

Advances in Knowledge 䡲 Longitudinal MR perfusion imaging of conservatively treated lowgrade gliomas showed differences in serial relative cerebral blood volume (rCBV) measurements between tumors that underwent malignant transformation and those that remained stable during a defined study period. 䡲 Transformers showed a marked increase in rCBV that could be observed up to 12, and in some instances, 18 months prior to clinical and imaging transformation, whereas nontransformers had relatively stable rCBV measurements. 䡲 The rate of rCBV change, determined from two studies 6 months apart, can help distinguish transformers from nontransformers as early as the interval 18 –12 months prior to transformation. Radiology: Volume 247: Number 1—April 2008

tion, and the development of areas of focal contrast enhancement is the most commonly used sign of tumor progression in clinical practice; this has proved a more reliable indicator of malignancy in gliomas than border definition, mass effect, necrosis, and hemorrhage (3–5). However, up to one-third of malignant gliomas do not enhance (6), and certain subtypes of low-grade gliomas show enhancement (typically gangliogliomas and pilocytic astrocytomas and occasionally oligodendrogliomas [7]). Contrast enhancement alone is therefore a limited differentiator between high- and low-grade gliomas. Moreover, the point at which enhancement appears during malignant transformation in a preexisting low-grade lesion is uncertain. There is therefore a need for additional markers of malignant change in gliomas that ideally reflect the earliest stages of the transformation process. Vascular proliferation (angiogenesis) is an important histologic hallmark of malignancy in glial tumors. Pathologic contrast enhancement in tumors indicates local disruption of the blood-brain barrier; this is an indirect marker of angiogenesis, as the walls of the new vessels may be deficient and more permeable. There is evidence that magnetic resonance (MR) perfusion imaging is a sensitive marker of the microvascular density and histologic grade of gliomas. Relative cerebral blood volume (rCBV) measurements correlate closely with angiographic and histologic markers of tumor vascularity (8) and are more elevated in high-grade than in low-grade gliomas (8–14). A correlation between rCBV and the expression of vascular endothelial growth factor has been demonstrated by using immunohistochemical staining of surgical specimens (15). Vascular endothelial growth factor is an

Implication for Patient Care 䡲 rCBV determination can help identify patients with low-grade gliomas who are at high risk of early malignant transformation and who might benefit from early aggressive therapy.

Danchaivijitr et al

important determinant of angiogenesis and is a prognostic marker in gliomas (16). A recent study (17) of patients referred for preoperative assessment of low-grade glioma showed that baseline rCBV measurements obtained prior to surgery correlated inversely with the time to subsequent tumor progression. Thus, the purpose of our study was to prospectively perform longitudinal MR perfusion imaging of conservatively treated low-grade gliomas to determine whether rCBV changes precede malignant transformation as defined by conventional MR imaging and clinical criteria.

Materials and Methods Patients Twenty-one patients with low-grade glioma (WHO grade II) recruited from the neuro-oncology clinic of our institution between August 2000 and August 2002 gave informed consent to participate in a longitudinal multimodal MR imaging study approved by the local research ethics committee. All patients presented with seizures, which were controlled by anticonvulsant medication. No patient had a neurologic deficit, and only those with low-grade gliomas confirmed with stereotactic biopsy were included in the study (n ⫽ 13). The other eight patients

Published online 10.1148/radiol.2471062089 Radiology 2008; 247:170 –178 Abbreviations: rCBV ⫽ relative cerebral blood volume ROI ⫽ region of interest WHO ⫽ World Health Organization Author contributions: Guarantors of integrity of entire study, N.D., A.D.W., J.H.R., H.R.J.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, N.D., A.D.W., G.B.C., H.R.J.; clinical studies, N.D., A.D.W., J.H.R., H.R.J.; experimental studies, G.B.C., P.S.T.; statistical analysis, A.D.W., D.J.T.; and manuscript editing, N.D., A.D.W., D.J.T., C.E.B., P.S.T., J.H.R., H.R.J. Authors stated no financial relationship to disclose.

171

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

had lesions in eloquent brain regions, in which the risk of biopsy was considered to be unacceptably high. Because their lesions were not confirmed histologically, these patients were excluded from this study. MR imaging studies were performed at 6-month intervals in all patients or until criteria for progression to a malignant tumor (malignant transformation) were fulfilled. The criteria were defined clinically as the development of a new focal deficit or new symptoms of raised intracranial pressure (but not worsening seizure control alone) and at imaging as the development of one or more new areas of pathologic contrast enhancement or substantial progression of previously stable baseline enhancement (4,5). Volume change alone was not regarded as indicative of tumor progression. Of the 13 patients in this study, seven were men and six were women, with a mean age of 44.2 years (range, 29 – 69 years). All patients had histologically confirmed WHO grade II gliomas at study entry (eight astrocytomas, four oligodendrogliomas, and one oligoastrocytoma), and all patients completed imaging follow-up.

Data Acquisition MR imaging studies were performed with a clinical 1.5-T system (Signa LX; GE Medical Systems, Milwaukee, Wis). Perfusion-weighted images were acquired during the intravenous injection of a bolus of gadoterate meglumine (Dotarem) at a dose of 0.1 mmol per kilogram of body weight and a rate of 5 mL/sec by using a single-shot gradientecho echo-planar imaging sequence (repetition time msec/echo time msec, 1200/40; flip angle, 20°; field of view, 26 cm; matrix, 96 ⫻ 128; section thickness, 5 mm; intersection gap, 3 mm). A total of six sections were prescribed to give maximum possible coverage of the entire tumor volume. High-spatialresolution coronal T1-weighted images were acquired volumetrically by using a spoiled gradient-echo sequence (14.4/ 6.4; inversion time, 650 msec; matrix, 256 ⫻ 256; field of view, 24 ⫻ 24 cm; section thickness, 1.5 mm; contiguous 172

sections) before and after the administration of 0.2 mmol/kg gadoterate meglumine (an additional 0.1 mmol/kg was given intravenously after the perfusion study).

Data Analysis Dynamic susceptibility-weighted images were processed off line at a workstation (Advantage) by using proprietary image analysis software (FuncTool, version 1.9; GE Medical Systems). The beginning and end of the first-pass bolus were determined through inspection of time– signal intensity curves (N.D., A.D.W., and H.R.J., with 2, 7, and 9 years of experience in interpreting dynamic perfusion images, respectively), and care was taken to exclude any recirculationrelated signal. For rCBV calculations, only the area under the curve of the first-pass bolus was considered, as a simple measure to minimize the previously described (18) confounding effects of contrast agent leakage (Fig 1). Color-coded rCBV maps were generated and projected onto the T2weighted (b ⫽ 0 sec/mm2) images; this provided better visualization of tumor boundaries. Three neuroradiologists (N.D., A.D.W., and H.R.J.) placed at least six ROIs of 15–20 pixels within the tumor, on areas showing the highest intratumoral rCBV on the color-coded maps. Regions were agreed to by consensus, and care was taken not to include peritumoral arteries and veins in these ROIs; this was aided by simultaneous viewing of the source images acquired during maximum arterial and venous contrast agent concentrations. The maximum rCBV value in intratumoral ROIs was selected for quantitative analysis. This approach has been shown to provide the best interobserver and intraobserver reproducibility in previous studies (11,19). The rCBV values were expressed as ratios relative to those in an ROI in contralateral normalappearing white matter. Annual rates of change in rCBV were calculated by subtracting the rCBV in the preceding study from the rCBV at the relevant time point and normalizing to change over 12 months. The presence of pathologic contrast

Danchaivijitr et al

enhancement was established through visual inspection of T1-weighted images in consensus by three experienced neuroradiologists (N.D., A.D.W., and H.R.J, with 6, 10, and 15 years of experience in brain MR imaging, respectively).

Statistical Analysis The patients were classified as “transformers,” who showed clinical and/or radiologic tumor progression during the period of data collection for this study or “nontransformers,” whose disease remained stable clinically and at imaging during the period of data collection for this study and for at least 12 months afterward (12–36 months). A repeated-measures analysis of variance was performed for all time points to determine whether there were differences between the data as a whole. Student paired and unpaired t tests were subsequently used to determine whether there were significant differences in measured rCBV values between specific time points within each group and between the nontransformer and transformer groups at different time points; P ⬍ .05 was considered to indicate a significant difference. KaplanMeier survival curves were calculated on the basis of baseline (time of entry) rCBV values for each subject; a threshold value of 1.65 was chosen to provide maximum predictive value for our cohort. Statistical analysis was performed by using proprietary software (SPSS, version 11.5, SPSS, Chicago, Ill; and Intercooled STATA, version 9.2, Stata, College Station, Tex). Results Transformers and Nontransformers Seven of the 13 patients had tumors that transformed during the study period (Table 1). Histologic evidence of anaplastic change (WHO grade III) was available in five of these patients (Table 1); the remaining two had undergone biopsy at study entry (which revealed one astrocytoma and one oligodendroglioma) and subsequently showed unequivocal clinical and imaging signs of tumor progression. Repeat Radiology: Volume 247: Number 1—April 2008

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

biopsy was not performed in these patients because of refusal (n ⫽ 1) or the treating physician’s decision (n ⫽ 1). Time from study entry to transforma-

tion varied from 6 to 36 months (mean, 22.3 months). The condition of the remaining six patients remained stable clinically and at imaging over an

Figure 1

Figure 1: Sample set of susceptibility-weighted contrast material–enhanced perfusion MR imaging data used to calculate rCBV measurements. (a) Baseline MR image obtained with transverse gradient-echo echo-planar imaging sequence (1200/40) before arrival of contrast medium shows slightly hyperintense right frontal oligodendroglioma with central areas of hypointense calcification. (b) Gradient-echo echo-planar image at peak contrast medium concentration shows loss of signal intensity in anterior portion of the tumor and more marked loss of signal intensity in cerebral arteries and veins. Superimposed is intratumoral region of interest (ROI) used for rCBV calculation. Care was taken to exclude neighboring vessels from this ROI. (c) Graph shows time–signal intensity curve over intratumoral ROI. There is an initial large decrease in signal intensity during first pass of gadolinium bolus, followed by a smaller signal intensity decrease (recirculation) and subsequent elevation of the signal intensity above baseline because of T1 effects associated with contrast material leakage into the extracellular space. Vertical line ⫽ cutoff point for rCBV calculation, chosen to exclude recirculation and substantial contrast material leakage according to previously described methods (18). (d) Color overlay map, with first T2-weighted precontrast image as background and a color scale with a window setting of 11.600–44.800 (arbitrary values). To obtain the rCBV, the value in the intratumoral ROI was divided by the average value in three ROIs placed over normal-appearing white matter in the contralateral hemisphere (not shown). Radiology: Volume 247: Number 1—April 2008

Danchaivijitr et al

observation period of 12– 48 months (mean, 23 months). There was no significant difference (P ⫽ .463) between the age of transformers (mean, 41.4 years; range, 30 –51 years) and that of nontransformers (mean, 47.5 years; range, 26 – 69 years).

rCBV Measurements at Specific Time Points In transformers, rCBV was measured at study entry and at each 6-month follow-up examination. In nontransformers, rCBV data were also acquired at 6-month intervals. rCBV in this group was found not to change between study entry and 1 year; data were therefore included from study entry, from the examination performed 1 year later, and from the last available time point of follow-up (mean, 23 months from entry). In transformers, three perfusion studies in two patients were nondiagnostic for technical reasons (at 6 months before transformation in patient 12 and at 30 and 24 months before transformation in patient 13). Serial color-coded rCBV maps in nontransformers showed low rCBV within tumor margins, which remained stable with time (Fig 2). In transformers, localized regions of elevated rCBV evolved prior to the appearance of pathologic contrast enhancement (Fig 3). This was reflected in values of maximum rCBV for individual patients, which changed little in nontransformers (Fig 4a) but showed progressive increases in transformers up to the time point at which transformation criteria were met (Fig 4b). The analysis of variance of the various time points yielded a P value of .013, indicating that there were significant differences between the time points. This allowed tests between pairs of time points to be performed. The group maximum rCBVs in nontransformers were stable between study entry (1.31 ⫾ 0.63) and 1 year (1.37 ⫾ 0.61) and showed a slight increase of borderline significance at the last available follow-up study (1.52 ⫾ 0.78, P ⫽ .058) (Table 2). The initial rCBV at study entry was slightly higher in transformers than in 173

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

nontransformers, but this difference was not significant (P ⫽ .298) (Table 2). The mean rCBV in transformers at the point of malignant transformation was markedly elevated (5.36 ⫾ 3.01) compared with baseline measures in the same group (1.93 ⫾ 1.25, P ⫽ .004) and in the nontransformer group (1.31 ⫾ 0.63, P ⫽ .008) (Table 2). Furthermore, significantly higher rCBV measurements were observed 6 months (P ⫽ .049) and 12 months (P ⫽ .022) before transformation, and borderline significantly higher measurements were observed 18 months (P ⫽ .074) before transformation. Although there was some overlap between baseline measurements in nontransformers and those in transformers 18 months prior to transformation, the groups were well separated 12 and 6 months prior to transformation (Fig 5). Maximum rCBV measures also provided a predictor of transformation-free survival (Fig 6). Both groups contained oligodendrogliomas and oligoastrocytomas; the oligodendrogliomas, which had higher baseline rCBVs than the group as a whole, also showed a clear further rCBV increase as they approached transformation.

Danchaivijitr et al

Table 1 Patient Demographics, Length of Follow-up, and Histologic Results at Study Entry and at Point of Clinical and Radiologic Transformation to High-Grade Tumor Patient Group and Patient No./Sex/Age (y) Nontransformers (n ⫽ 6) 1/F/29 2/F/57 3/M/29 4/M/38 5/F/63 6/M/69 Transformers (n ⫽ 7) 7/F/47 8/M/35 9/M/51 10/M/48 11/F/42 12/F/30 13/M/37

No. of Studies

Duration of Follow-up (mo)

Histologic Finding at Initial Biopsy

Histologic Finding at Biopsy at Transformation Point

3 5 6 3 5 7

12 24 30 12 24 36

Grade II astrocytoma Grade II astrocytoma Grade II astrocytoma Grade II astrocytoma Grade II oligodendroglioma Grade II oligoastrocytoma

... ... ... ... ... ...

7 6 2 2 5 3* 5*

36 30 6 6 24 18 36

Grade II astrocytoma Grade II astrocytoma Grade II astrocytoma Grade II oligodendroglioma Grade II oligodendroglioma Grade II oligodendroglioma Grade II astrocytoma

Grade III astrocytoma Grade III astrocytoma Grade III astrocytoma Grade III oligoastrocytoma Grade III oligodendroglioma ... ...

* Some perfusion studies failed technically (see Results); this explains discrepancies between length of follow-up and number of studies available for analysis.

Figure 2

Changes in rCBV over Time A difference between nontransformers and transformers was observed in the rate of change in rCBV. The mean increase in rCBV was 0.06 per year ⫾ 0.12 at the 1-year study in the nontransformers, compared with 1.1 per year ⫾ 0.8 (P ⫽ .043) in the period 18 –12 months prior to transformation in transformers (Table 3). It was not significantly different from that in nontransformers in the period 12– 6 months prior to transformation (2.52 ⫾ 2.47, P ⫽ .140) but was significant again (4.18 ⫾ 2.24, P ⫽ .006) in the period 6 months before transformation (Table 3). Discussion To our knowledge, ours is the first longitudinal study of MR perfusion imaging in conservatively treated supratentorial gliomas in adults. 174

Figure 2: Patient 3. Images from serial MR perfusion study in 29-year-old man with left frontal low-grade astrocytoma that did not undergo malignant transformation during an observation period of 30 months. The transverse rCBV color overlay maps (obtained by using the more T2-weighted first image of the perfusion series as background) are windowed to show areas with greater rCBV than white matter. (a) rCBV map at study entry shows tumor with low rCBV (maximum measured rCBV, 0.76). (b) After 30 months, there has been some increase in tumor volume, but the maximum rCBV remains low (0.96). Radiology: Volume 247: Number 1—April 2008

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

A number of cross-sectional studies have revealed the usefulness of MR perfusion imaging for preoperative grading of gliomas (8–14) and have indicated that the technique significantly augments the sensitivity and specificity of

conventional MR imaging in predicting histologic grade (11). The optimal perfusion imaging technique in this context remains a contentious issue. Spin-echo images are more sensitive to signal change in the micro-

Danchaivijitr et al

vasculature than in large vessels (20); this is viewed as an advantage by some investigators because it minimizes signal contamination from larger vessels at the periphery of the tumor (12). Gradient-echo techniques, however, have

Figure 3

Figure 3: Images from serial MR perfusion studies in 30-year-old patient with oligodendroglioma that showed progression to high-grade tumor 18 months after study entry. (a– c) Transverse rCBV overlay images. (d–f) Transverse reformatted images obtained with double-dose contrast-enhanced volumetric T1-weighted sequence (spoiled gradient echo; 14.4/6.4; inversion time, 650 msec). (a) rCBV map at baseline shows small area of elevated rCBV (arrow) measuring up to 4.52. (b) rCBV map 6 months before transformation shows larger area of increased rCBV (arrow) measuring up to 8.32. (c) rCBV map at transformation shows further increase in area with elevated rCBV (arrow), which now reaches a maximum value of 12.04. (d) Baseline contrast-enhanced T1-weighted image shows a hypointense tumor without pathologic enhancement. (e) Six months before transformation, there is no evidence of pathologic intratumoral enhancement, despite a markedly increased rCBV. (f) At transformation, there is irregular enhancement in center of the tumor (arrow); the area of pathologic enhancement is much smaller than the region of increased rCBV. Radiology: Volume 247: Number 1—April 2008

175

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

Danchaivijitr et al

Figure 4

Figure 4: (a) Graph shows rCBV measurements in nontransformers at study entry, at 1 year, and at last available follow-up point; the longest follow-up was 48 months. rCBV remains virtually unchanged at 1 year, and there is only a slight increase in rCBV at the last follow-up point in patient 5 that is much less marked than that in the transformers (Fig 2b). The highest rCBV measurement in this group was in patient 5, who had an oligodendroglioma. (b) Graph shows serial rCBV measurements in transformers. The last measurement point corresponds to the point of transformation. The preceding measurements are counted backward from the point of transformation, and the time interval between each measurement point is 6 months. Most transformers showed a marked increase in rCBV starting 12 months (and, in some patients, even 18 months) before transformation. The patient with the highest rCBV measurement in this group (patient 11) also had an oligodendroglioma.

been shown to be better than spin-echo techniques for discriminating low-grade from high-grade gliomas (10,13). This may be explained by a greater sensitivity of gradient-echo images to larger intratumoral vessels in high-grade tumors. We used a gradient-echo technique and took care to avoid any large peritumoral vessels by superimposing the ROIs on source images that showed maximum signal intensity decrease in large vessels. Using conventional criteria of new contrast enhancement (3–5) and clinical deterioration as indicators of malignant progression, we found striking differences in rCBVs between nontransformers and transformers— both in values at individual time points and in their rate of change over time. In nontransformers, rCBV measurements remained low throughout the study period (mean, 23 months). The mean rCBV measurements in this group accord with previously published data for low-grade gliomas. Law et al (11) found that an rCBV 176

Table 2 rCBV Values in Transformers and Nontransformers at Various Time Points Patient Group and Time Point Nontransformers Baseline (n ⫽ 6) 1 Year (n ⫽ 6) Last point of follow-up (n ⫽ 6) Transformers Baseline (n ⫽ 7) 18 Months before transformation (n ⫽ 5) 12 Months before transformation (n ⫽ 5) 6 Months before transformation (n ⫽ 6) Transformation (n ⫽ 7)

Mean ⫾ Standard Deviation

Range

1.31 ⫾ 0.63 1.37 ⫾ 0.61 1.52 ⫾ 0.78

0.70–2.29 0.79–2.28 0.79–2.81

1.93 ⫾ 1.25 2.59 ⫾ 1.40 3.14 ⫾ 1.48* 3.65 ⫾ 2.48* 5.36 ⫾ 3.01*

0.88–4.52 1.09–4.80 1.11–5.28 1.69–8.38 3.39–12.04

* Significantly different from value at baseline. (For P values, see Results section of text.)

threshold value of 1.75 yielded a sensitivity of 95% and a positive predictive value of 87% for distinguishing between low- and high-grade gliomas. In a more recent study (17) of a cohort of lowgrade gliomas (including 21 astrocytomas and 14 oligodendrogliomas), the same research group showed that rCBV

values greater than 1.75 were associated with more rapid tumor progression. The variations between threshold rCBV values chosen in different studies are small and are easily accounted for by differences in technique, patient group, and analysis software. Only one patient in our nontransRadiology: Volume 247: Number 1—April 2008

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

Figure 5

Figure 5: Box plot shows yearly rate of change in rCBV for transformers and nontransformers. In the transformers, the rate of change was calculated from two successive studies performed at 6 month intervals. In the nontransformers, the rate of change was calculated from the time of last available follow-up to baseline, thereby measuring the largest possible rate of change in the nontransformers during the study period. These measurements were adjusted and expressed as yearly rates of change in both groups. There was a marked and significant difference in the yearly rate of change of rCBV between nontransformers and transformers from 12 months before transformation onward. NTBaseline ⫽ nontransformers at baseline, T-18m ⫽ transformation minus 18 months, T-12m ⫽ transformation minus 12 months, T-6m ⫽ transformation minus 6 months, T ⫽ transformation.

Figure 6

former group had a baseline rCBV higher than 1.75; this patient had a grade II oligodendroglioma with an rCBV of 2.29. Low-grade oligodendroglial tumors tend to have higher rCBVs than do grade II astrocytomas; this is a potential confounder in grading glial tumors by using perfusion criteria alone (12,21). It is important to note that the initial MR imaging study is a somewhat arbitrary time point in the natural history of a low-grade glioma. The timing of clinical presentation and, hence, the initial imaging investigation depends on when the patient becomes clinically symptomatic, which does not seem to depend on tumor size or location. The interval between the initial study and transformation in transformers was highly variable (6 –36 months) and therefore cannot also provide a meaningful baseline group rCBV measure. The average baseline rCBV measurement in transformers was slightly higher than that in nontransformers, although this did not reach statistical significance. The transformer group data were influenced by two patients with high rCBVs who fulfilled transformation criteria after only 6 months and who are likely to have undergone early focal transformation that was not detected with biopsy and by the greater number of oligodendrogliomas in the transformer group. It is therefore more relevant to compare rCBV in transformers at different time points leading up to transformation with early measurements in nontransformers (a

Danchaivijitr et al

group in which malignant transformation is known not to be imminent). We believe the small within-subject variation in rCBV in the nontransformers in our study is an indicator of the longitudinal stability of our method and reflects the absence of marked angiogenic activity in this group. Mean rCBV at the point of transformation was 5.36 ⫾ 3.01, which is in broad agreement with published data in high-grade gliomas: Sugahara et al (13) found mean maximum rCBV values of 5.84 in anaplastic astrocytomas and 7.32 in glioblastomas, and Yang et al (14) found values of 6.1 for high-grade (mixed grade III and IV) gliomas. In transformers, we found not only a marked difference in rCBV at the time of transformation but also a steady increase in rCBV to a degree that was detectable up to 18 months beforehand (although not statistically significant at P ⫽ .074). Compared with nontransformers, in transformers, the mean rCBV was significantly higher 12 months (P ⫽ .022) and 6 months (P ⫽ .049) before contrast enhancement became apparent on T1weighted MR images. This indicates that an increase in microvascular density occurs well in advance of bloodbrain barrier leakage reflected by pathologic contrast enhancement. The mean rCBVs in our transformer group 6 and 12 months before enhancement were 3.14 and 3.65, respectively. This is in concordance with results of a

Table 3 Yearly Rates of Change in rCBV in Nontransformers and Transformers as Calculated from Two Successive rCBV Measurements Patient Group and Time Point

Figure 6: Kaplan-Meier survival curve for progression-free survival for tumors with rCBV of less than 1.65 versus that for tumors with rCBV of 1.65 or greater, on the basis of a single measurement. P ⬍ .004, log-rank test for equality of survivor functions. Radiology: Volume 247: Number 1—April 2008

Nontransformers Baseline to 1 year (n ⫽ 6) 1 Year to last point of follow-up (n ⫽ 4) Transformers 18 Months before transformation to 12 months before transformation (n ⫽ 5) 12 Months before transformation to 6 months before transformation (n ⫽ 4) 6 Months before transformation to transformation (n ⫽ 6)

Rate of Change per Year* Mean ⫾ SD Range 0.06 ⫾ 0.12 0.19 ⫾ 0.23

⫺0.14 to 0.21 ⫺0.14 to 0.26

1.1 ⫾ 0.8

0.05 to 1.92

2.52 ⫾ 2.47

1.1 to 6.2

4.18 ⫾ 2.24

1.29 to 7.33

* Calculated from differences in rCBV from examinations 6 months apart.

177

NEURORADIOLOGY: rCBV Measurements in Low-Grade Gliomas

cross-sectional study of rCBV, vascular endothelial growth factor expression, and tumor histologic features in 20 nonenhancing gliomas (15); all seven high-grade tumors in that study had rCBVs ranging from 2.8 to 3.7 (mean, 2.75)—values significantly higher than those of the 13 low-grade tumors. Our data support the likelihood that the cellular processes underlying malignant transformation may occur 12 months or more before contrast enhancement is detectable. The time at which histologic progression to highgrade tumor occurred in our patients remains uncertain, as we did not perform interim biopsies between baseline and the point of contrast enhancement. This needs to be addressed by a further longitudinal study in which the decision to perform repeat biopsies is based on changes in rCBV measurements rather than on traditional criteria for malignant progression, such as contrast enhancement. In our relatively small series, both absolute rCBV values and annual rates of change (calculated from two studies as few as 6 months apart), discriminated between low-grade gliomas that did not transform and tumors that were within 12 months of transformation as defined by conventional criteria. These are therefore likely to be useful parameters in clinical practice for stratifying those tumors at high risk of early malignant progression. Rate-of-change measures may be particularly helpful in oligodendrogliomas. In concordance with previous investigators (12,21), we found low-grade oligodendrogliomas to have higher rCBVs than astrocytomas; this contributes to overlap of single-point measurements between the transformer and nontransformer groups and may confound assessment of tumor grade. The nontransforming oligodendroglioma and oligoastrocytoma showed little change in rCBV over time, whereas the rCBV in transforming oligodendrogliomas increased considerably (from 4 to 12, 3.5 to 8, and 2.5 to 4). Our study was limited by the relatively small numbers of patients. Some

178

selection bias, related to patients who elected to be treated conservatively, may also have occurred. The number of sections from which perfusion measurements could be made was also a potential limitation, as coverage of the entire tumor may have been incomplete. There was also uncertainty in tumor grade at presentation due to sampling limitations inherent in stereotactic biopsy, the lack of interval histologic data, and, in two patients, the lack of final histologic data. In conclusion, we have demonstrated that increases in rCBV precede the development of contrast enhancement by at least 12 months in transforming low-grade gliomas; rCBV increase is therefore likely to provide an earlier noninvasive indicator of malignant progression. This has important implications for clinical management, by helping to identify patients most likely to benefit from early treatment. We therefore recommend that MR perfusion imaging be used routinely in the initial assessment and subsequent evaluation of patients with low-grade gliomas who are treated conservatively.

References 1. Johannesen TB, Langmark F, Lote K. Progress in long-term survival in adult patients with supratentorial low-grade gliomas: a population-based study of 993 patients in whom tumors were diagnosed between 1970 and 1993. J Neurosurg 2003; 99(5):854 – 862. 2. van den Bent MJ, Afra D, de Witte O, et al. Long-term efficacy of early versus delayed radiotherapy for low-grade astrocytoma and oligodendroglioma in adults: the EORTC 22845 randomised trial. Lancet 2005;366(9490):985–990. 3. Daumas-Duport C, Tucker ML, Kolles H, et al. Oligodendrogliomas. II. A new grading system based on morphological and imaging criteria. J Neurooncol 1997;34(1):61–78. 4. Mihara F, Numaguchi Y, Rothman M, Sato S, Fiandaca MS. MR imaging of adult supratentorial astrocytomas: an attempt of semi-automatic grading. Radiat Med 1995;13(1):5–9. 5. Pierallini A, Bonamini M, Bozzao A, et al. Supratentorial diffuse astrocytic tumours: proposal of an MRI classification. Eur Radiol 1997;7(3):395–399. 6. Scott JN, Brasher PM, Sevick RJ, Rewcastle NB, Forsyth PA. How often are nonenhancing supratentorial gliomas malignant? a population study. Neurology 2002;59(6):947–949. 7. White ML, Zhang Y, Kirby P, Ryken TC. Can tumor contrast enhancement be used as a criterion for differentiating tumor grades of oligodendrogliomas? AJNR Am J Neuroradiol 2005;26(4): 784 –790.

Danchaivijitr et al

8. Sugahara T, Korogi Y, Kochi M, et al. Correlation of MR imaging-determined cerebral blood maps with histologic and angiographic determination of vascularity of gliomas. AJR Am J Roentgenol 1998;171(6):1479 –1486. 9. Aronen HJ, Gazit IE, Louis DN, et al. Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology 1994;191(1):41– 45. 10. Donahue KM, Krouwer HG, Rand SD, et al. Utility of simultaneously acquired gradient-echo and spin-echo cerebral blood volume and morphology maps in brain tumor patients. Magn Reson Med 2000;43(6):845– 853. 11. Law M, Yang S, Wang H, et al. Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol 2003;24(10): 1989 –1998. 12. Lev MH, Ozsunar Y, Henson JW, et al. Glial tumor grading and outcome prediction using dynamic spin-echo MR susceptibility mapping compared with conventional contrast-enhanced MR: confounding effect of elevated rCBV of oligodendrogliomas [corrected]. AJNR Am J Neuroradiol 2004;25(2):214 –221. [Published correction appears in AJNR Am J Neuroradiol 2004;25(3):B1.] 13. Sugahara T, Korogi Y, Kochi M, Ushio Y, Takahashi M. Perfusion-sensitive MR imaging of gliomas: comparison between gradient-echo and spin-echo echo-planar imaging techniques. AJNR Am J Neuroradiol 2001;22(7):1306 –1315. 14. Yang D, Korogi Y, Sugahara T, et al. Cerebral gliomas: prospective comparison of multivoxel 2D chemical-shift imaging proton MR spectroscopy, echoplanar perfusion and diffusion-weighted MRI. Neuroradiology 2002;44(8):656 – 666. 15. Maia AC Jr, Malheiros SM, da Rocha AJ, et al. MR cerebral blood volume maps correlated with vascular endothelial growth factor expression and tumor grade in nonenhancing gliomas. AJNR Am J Neuroradiol 2005;26(4):777–783. 16. Abdulrauf SI, Edvardsen K, Ho KL, Yang XY, Rock JP, Rosenblum ML. Vascular endothelial growth factor expression and vascular density as prognostic markers of survival in patients with low-grade astrocytoma. J Neurosurg 1998;88(3): 513–520. 17. Law M, Oh S, Babb JS, et al. Low-grade gliomas: dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging—prediction of patient clinical response. Radiology 2006;238(2):658 – 667. 18. Cha S, Knopp EA, Johnson G, Wetzel SG, Litt AW, Zagzag D. Intracranial mass lesions: dynamic contrast-enhanced susceptibility-weighted echo-planar perfusion MR imaging. Radiology 2002;223(1):11–29. 19. Wetzel SG, Cha S, Johnson G, et al. Relative cerebral blood volume measurements in intracranial mass lesions: interobserver and intraobserver reproducibility study. Radiology 2002; 224(3):797– 803. 20. Kennan R, Ja¨ger HR. T2 and T2*-w DCE-MRI: blood perfusion and volume estimation using bolus tracking. In: Tofts P, ed. Quantitative MRI of the brain. Chichester, England: Wiley, 2003; 365–412. 21. Cha S, Tihan T, Crawford F, et al. Differentiation of low-grade oligodendrogliomas from low-grade astrocytomas by using quantitative blood-volume measurements derived from dynamic susceptibility contrast-enhanced MR imaging. AJNR Am J Neuroradiol 2005;26(2):266 –273.

Radiology: Volume 247: Number 1—April 2008

Radiology 2008 This is your reprint order form or pro forma invoice (Please keep a copy of this document for your records.)

Reprint order forms and purchase orders or prepayments must be received 72 hours after receipt of form either by mail or by fax at 410-820-9765. It is the policy of Cadmus Reprints to issue one invoice per order. Please print clearly. Author Name _______________________________________________________________________________________________ Title of Article _______________________________________________________________________________________________ Issue of Journal_______________________________ Reprint # _____________ Publication Date ________________ Number of Pages_______________________________ KB # _____________ Symbol Radiology Color in Article? Yes / No (Please Circle) Please include the journal name and reprint number or manuscript number on your purchase order or other correspondence.

Order and Shipping Information Reprint Costs (Please see page 2 of 2 for reprint costs/fees.) ________ Number of reprints ordered

Shipping Address (cannot ship to a P.O. Box) Please Print Clearly

$_________

________ Number of color reprints ordered $_________ ________ Number of covers ordered

$_________

Subtotal $_________ Taxes

$_________

(Add appropriate sales tax for Virginia, Maryland, Pennsylvania, and the District of Columbia or Canadian GST to the reprints if your order is to be shipped to these locations.)

First address included, add $32 for each additional shipping address

TOTAL

$_________

$_________

Name ___________________________________________ Institution _________________________________________ Street ___________________________________________ City ____________________ State _____ Zip ___________ Country ___________________________________________ Quantity___________________ Fax ___________________ Phone: Day _________________ Evening _______________ E-mail Address _____________________________________

Additional Shipping Address* (cannot ship to a P.O. Box) Name ___________________________________________ Institution _________________________________________ Street ___________________________________________ City ________________ State ______ Zip ___________ Country _________________________________________ Quantity __________________ Fax __________________ Phone: Day ________________ Evening ______________ E-mail Address ____________________________________ * Add $32 for each additional shipping address

Payment and Credit Card Details

Invoice or Credit Card Information

Enclosed: Personal Check ___________ Credit Card Payment Details _________

Invoice Address

Please Print Clearly

Please complete Invoice address as it appears on credit card statement

Checks must be paid in U.S. dollars and drawn on a U.S. Bank.

Credit Card: __ VISA __ Am. Exp. __ MasterCard Card Number __________________________________ Expiration Date_________________________________ Signature: _____________________________________ Please send your order form and prepayment made payable to:

Cadmus Reprints P.O. Box 751903 Charlotte, NC 28275-1903

Name ____________________________________________ Institution ________________________________________ Department _______________________________________ Street ____________________________________________ City ________________________ State _____ Zip _______ Country ___________________________________________ Phone _____________________ Fax _________________ E-mail Address _____________________________________ Cadmus will process credit cards and Cadmus Journal Services will appear on the credit card statement.

Note: Do not send express packages to this location, PO Box. FEIN #:541274108

If you don’t mail your order form, you may fax it to 410-820-9765 with your credit card information.

Signature __________________________________________ Date _______________________________________ Signature is required. By signing this form, the author agrees to accept the responsibility for the payment of reprints and/or all charges described in this document. RB-9/26/07

Page 1 of 2

Radiology 2008 Black and White Reprint Prices

Color Reprint Prices

Domestic (USA only) # of Pages 1-4 5-8 9-12 13-16 17-20 21-24 25-28 29-32 Covers

50 $221 $355 $466 $576 $694 $809 $928 $1,042 $97

100

200

300

$233 $268 $285 $382 $432 $466 $513 $595 $652 $640 $749 $830 $775 $906 $1,017 $906 $1,071 $1,200 $1,041 $1,242 $1,390 $1,178 $1,403 $1,568 $118 $215 $323

Domestic (USA only) 400

500

$303 $510 $714 $912 $1,117 $1,321 $1,544 $1,751 $442

$323 $544 $775 $995 $1,220 $1,471 $1,688 $1,924 $555

# of Pages 1-4 5-8 9-12 13-16 17-20 21-24 25-28 29-32 Covers

International (includes Canada and Mexico) # of Pages 1-4 5-8 9-12 13-16 17-20 21-24 25-28 29-32 Covers

50

100

200

300

400

500

$223 $349 $486 $615 $759 $897 $1,033 $1,175 $97

$239 $401 $517 $651 $787 $924 $1,071 $1,208 $118

$352 $601 $852 $1,105 $1,357 $1,611 $1,873 $2,122 $215

$473 $849 $1,232 $1,609 $1,997 $2,376 $2,757 $3,138 $323

$597 $1,099 $1,609 $2,117 $2,626 $3,135 $3,650 $4,162 $442

$719 $1,349 $1,992 $2,624 $3,260 $3,905 $4,536 $5,180 $555

International (includes Canada and Mexico))

50

100

200

300

400

500

$272 $428 $580 $724 $878 $1,022 $1,176 $1,316 $156

$283 $455 $626 $786 $958 $1,119 $1,291 $1,452 $176

$340 $576 $805 $1,023 $1,246 $1,474 $1,700 $1,936 $335

$397 $675 $964 $1,232 $1,520 $1,795 $2,070 $2,355 $525

$446 $784 $1,115 $1,445 $1,774 $2,108 $2,450 $2,784 $716

$506 $884 $1,278 $1,652 $2,030 $2,426 $2,813 $3,209 $905

Minimum order is 50 copies. For orders larger than 500 copies, please consult Cadmus Reprints at 800-407-9190.

Reprint Cover Cover prices are listed above. The cover will include the publication title, article title, and author name in black.

# of Pages 1-4 5-8 9-12 13-16 17-20 21-24 25-28 29-32 Covers

50

100

200

300

400

500

$278 $429 $604 $766 $945 $1,110 $1,290 $1,455 $156

$290 $472 $629 $797 $972 $1,139 $1,321 $1,482 $176

$424 $746 $1,061 $1,378 $1,698 $2,015 $2,333 $2,652 $335

$586 $1,058 $1,545 $2,013 $2,499 $2,970 $3,437 $3,924 $525

$741 $1,374 $2,011 $2,647 $3,282 $3,921 $4,556 $5,193 $716

$904 $1,690 $2,494 $3,280 $4,069 $4,873 $5,661 $6,462 $905

Tax Due Residents of Virginia, Maryland, Pennsylvania, and the District of Columbia are required to add the appropriate sales tax to each reprint order. For orders shipped to Canada, please add 7% Canadian GST unless exemption is claimed.

Ordering

Shipping Shipping costs are included in the reprint prices. Domestic orders are shipped via UPS Ground service. Foreign orders are shipped via a proof of delivery air service.

Reprint order forms and purchase order or prepayment is required to process your order. Please reference journal name and reprint number or manuscript number on any correspondence. You may use the reverse side of this form as a proforma invoice. Please return your order form and prepayment to:

Multiple Shipments

Cadmus Reprints P.O. Box 751903 Charlotte, NC 28275-1903

Orders can be shipped to more than one location. Please be aware that it will cost $32 for each additional location.

Delivery Your order will be shipped within 2 weeks of the journal print date. Allow extra time for delivery.

Note: Do not send express packages to this location, PO Box. FEIN #:541274108 Please direct all inquiries to: Rose A. Baynard 800-407-9190 (toll free number) 410-819-3966 (direct number) 410-820-9765 (FAX number) [email protected] (e-mail)

Page 2 of 2

Reprint Order Forms and purchase order or prepayments must be received 72 hours after receipt of form.