J Neurosurg 121:367–373, 2014 ©AANS, 2014
Prediction of methylguanine methyltransferase promoter methylation in glioblastoma using dynamic contrast-enhanced magnetic resonance and diffusion tensor imaging Clinical article Sung Soo Ahn, M.D., Ph.D.,1 Na-Young Shin, M.D.,1 Jong Hee Chang, M.D., Ph.D., 2 Se Hoon Kim, M.D., Ph.D., 3 Eui Hyun Kim, M.D., 2 Dong Wook Kim, Ph.D., 4 and Seung-Koo Lee, M.D., Ph.D.1 Department of Radiology, Research Institute of Radiological Science, Yonsei University College of Medicine; Departments of 2Neurosurgery and 3Pathology, Brain Research Institute, Yonsei University College of Medicine; and 4Biostatistics Collaboration Unit, Yonsei University College of Medicine, Seoul, Korea
1
Object. The methylation status of the methylguanine methyltransferase (MGMT) promoter has been associated with treatment response in glioblastoma. The authors aimed to assess whether MGMT methylation status can be predicted by dynamic contrast-enhanced (DCE) MRI and diffusion tensor imaging (DTI). Methods. This retrospective study included 43 patients with pathologically diagnosed glioblastoma who had undergone preoperative DCE-MRI and DTI and whose MGMT methylation status was available. The imaging features were qualitatively assessed using conventional MR images. Regions of interest analyses for DCE-MRI permeability parameters (transfer constant [Ktrans], rate transfer coefficient [Kep], and volume fraction of extravascular extracellular space [Ve]) and DTI parameters (apparent diffusion coefficient [ADC] and fractional anisotropy [FA]) were performed on the enhancing solid portion of the glioblastoma. Chi-square or Mann-Whitney tests were used to evaluate relationships between MGMT methylation and imaging parameters. The authors performed receiver operating characteristic curve analysis to find the optimal cutoff value for the presence of MGMT methylation. Results. MGMT methylation was not significantly associated with any imaging features on conventional MR images. Ktrans values were significantly higher in the MGMT methylated group (median 0.091 vs 0.053 min-1, p = 0.018). However, Kep, Ve, ADC, and FA were not significantly different between the 2 groups. The optimal cutoff value for the presence of MGMT methylation was Ktrans > 0.086 min-1 with an area under the curve of 0.756, a sensitivity of 56.3%, and a specificity of 85.2%. Conclusions. Ktrans may serve as a potential imaging biomarker to predict MGMT methylation status preoperatively in glioblastoma; however, further investigation with a larger cohort is necessary. (http://thejns.org/doi/abs/10.3171/2014.5.JNS132279)
Key Words • glioblastoma • O-6-methylguanine-DNA methyltransferase • magnetic resonance imaging • permeability • diffusion tensor imaging • oncology
M
ethylguanine methyltransferase (MGMT) is a key enzyme in DNA repair because it removes alkyl groups from the O6 position of guanine, an important site of DNA alkylation.7 Epigenetic silencing of the MGMT gene by promoter methylation is associated with low levels of MGMT proteins and consequently
Abbreviations used in this paper: ADC = apparent diffusion coefficient; BBB = blood-brain barrier; DCE = dynamic contrastenhanced; DTI = diffusion tensor imaging; FA = fractional anisotropy; Kep = rate transfer coefficient; Ktrans = transfer constant; MGMT = methylguanine methyltransferase; ROI = region of interest; Ve = volume fraction of extravascular extracellular space.
J Neurosurg / Volume 121 / August 2014
reduces DNA repair activity against alkylating agents.16,19 Therefore, glioblastoma with MGMT promoter methylation can be expected to have a more favorable response to temozolomide, a DNA alkylating agent. Several studies have reported that the methylation status of the MGMT gene promoter is associated with treatment response to chemotherapy and prognosis.10,11 MGMT methylation has been reported in 30%–60% of glioblastomas.19 Although genetic analysis using surgical specimens is the This article contains some figures that are displayed in color online but in black-and-white in the print edition.
367
S. S. Ahn et al. gold standard for assessing the methylation status of MGMT, there are limitations to this process. It requires an invasive procedure where there is still the possibility of incomplete biopsy sampling due to a spatially heterogeneous glioblastoma. Sufficient samples may be difficult to obtain if the tumor is inaccessible to the surgeon. Therefore, it would be helpful in predicting treatment response and prognosis if it were possible to noninvasively determine MGMT methylation status with preoperative imaging. A few studies have described distinctive radiological features of glioblastoma with MGMT methylation using conventional MRI and diffusion-weighted or diffusion tensor imaging (DTI).5,14,15 However, there are some discrepancies between the reports; some investigators have reported that the apparent diffusion coefficient (ADC) value is lower in the MGMT methylated group,15 whereas others have reported that the ADC ratio is significantly higher in the MGMT methylated group.14 While some studies have reported an enhancement pattern or tumor margin characteristic that appears to be associated with MGMT methylation status, others have not found such an association.5,6,9,14 Dynamic contrast-enhanced (DCE) MRI using a contrast agent is an emerging MRI technique based on kinetic modeling of microvascular permeability that enables quantification of blood-brain barrier (BBB) breakdown.18 Several studies have shown efficacy of permeability parameters from DCE-MRI in glioma grading.12,17 However, no study has evaluated the association between permeability parameters and MGMT methylation status in glioblastoma. Therefore, the aim of this study was to identify imaging biomarkers for MGMT methylation using quantitative parameters from DTI and DCE-MRI in addition to conventional imaging features.
Methods
The institutional review board approved this retrospective study and did not require patient approval or informed consent for the review of patient images. However, we did obtain informed consent for MGMT promoter gene evaluation at the time of surgery. Patients
We retrospectively studied 49 patients between October 2011 and January 2013 from the neurooncology database at our institution on the basis of the following criteria: 1) a histopathological diagnosis of glioblastoma based on the WHO grading system and 2) a record of preoperative MRI performed with DTI and DCE-MRI. We excluded 6 patients due to the following conditions: inadequate MRI (n = 3), failure to obtain informed consent for MGMT promoter gene evaluation (n = 2), or stereotactic biopsy (n = 1). Accordingly, 43 patients were enrolled (25 men and 18 women; mean age 58 ± 14.5 years). The characteristics of the 43 patients are shown in Table 1. The mean interval between preoperative MRI and surgery was 5.8 ± 4.3 days. Of the 43 patients, 18 underwent gross-total resection, 21 underwent subtotal resection, and 4 underwent partial resection. All pathological specimens were examined by an experienced neuropathologist (K.S.H.), and the MGMT methylation status
368
TABLE 1: Characteristics of 43 patients with glioblastoma* Parameter mean age sex M F extent of resection total subtotal partial
Methylated Unmethylated MGMT (n = 16) MGMT (n = 27)
Total (n = 43)
64.3 ± 9.6
54.2 ± 15.7
58 ± 14.5
9 (56.2) 7 (43.8)
16 (59.3) 11 (40.7)
25 (58.1) 18 (41.9)
8 (50) 7 (43.8) 1 (6.3)
10 (37) 14 (51.9) 3 (11.1)
18 (41.9) 21 (48.8) 4 (9.3)
* Data are presented as numbers of patients (%) unless noted otherwise. The mean value is presented as ± SD.
was assessed with methylation-specific polymerase chain reaction. Image Acquisition
Preoperative MR images were obtained using a 3.0-T system (Achieva, Philips) and an 8-channel SENSE head coil. The preoperative evaluation MRI protocol included the following conventional sequences: pre- and postcontrast T1-weighted imaging (TR 2000 msec, TE 10 msec, FOV 240 mm, slice thickness 5 mm, and matrix 256 × 256), T2-weighted imaging (TR 3000 msec, TE 80 msec, FOV 240 mm, slice thickness 5 mm, and matrix 256 × 256), and fluid-attenuated inversion recovery (TR 10,000 msec, TE 125 msec, FOV 240 mm, slice thickness 5 mm, and matrix 256 × 256). For DCE-MRI, precontrast 3D T1-weighted images were obtained with the following parameters: TR 6.3 msec, TE 3.1 msec, FOV 240 mm, matrix 192 × 192 mm, slice thickness 3 mm, and flip angle 5°. After the precontrast scan, 60 DCE T1-wighted images were obtained with the same MR parameters except for an increased flip angle of 15°. After acquisition of the fifth image volume, gadolinium-based contrast (0.1 ml/kg gadobutrol, Gadovist, Bayer Schering Pharma) was injected at a rate of 3 ml/sec. The acquisition time for DCE-MRI was 378 seconds. Diffusion tensor images were obtained by applying 6 different directions of orthogonal diffusion gradients and b values of 1000 seconds/mm2 and 0 seconds/mm2 (TR 8400 msec, TE 77 msec, FOV 240 mm, slice thickness 2 mm, and matrix 128 × 128). Qualitative Imaging Analysis
Two neuroradiologists (S.S.A. and N.Y.S.) who were blinded to the patients’ molecular and clinical data reviewed the conventional MR images on a standard picture archiving and communication system. All tumors were assessed for the following imaging features: enhancing tumor margin (well or poorly defined); enhancement patterns (ring, nodular, or mixed enhancement); presence of edema, cysts, necrosis, and nonenhancing tumor; and heterogeneity of the signal intensity on the T2-weighted images. Briefly, necrosis was defined as regions of peJ Neurosurg / Volume 121 / August 2014
Prediction of MGMT methylation in glioblastoma using MRI ripheral and irregular enhancement surrounding areas of T2 hyperintensity. Nonenhancing tumor regions were defined as areas of intermediate T2 signal intensity associated with a mass effect and architectural distortion. Discordant interpretations were resolved by consensus. Quantitative Imaging Analysis
An experienced neuroradiologist (S.S.A.) who was blinded to molecular and clinical data performed postprocessing. Permeability maps from DCE-MRI, including volume transfer constant (Ktrans), rate transfer coefficient (Kep), and volume fraction of extravascular extracellular space (Ve) maps, were generated by off-line Pride tools, based on the pharmacokinetic model of Tofts and Kermode,18 provided by Philips Medical Systems. Postprocessing comprised motion correction of pixels from dynamic images, T1 mapping using different flip angles (5° and 15°), registration of pixels on a T1 map, vascular input function estimation, and pharmacokinetic modeling. All of these processes were automatically performed by Pride tools except for the drawing of regions of interest (ROIs) for vascular input function. The ROIs for the vascular input function were drawn from the vertical part of the superior sagittal sinus on the middle section of the scanned volume. For volume-based analysis, ROIs were drawn to contain all enhancing components of the tumor in each section of the Ktrans, Kep, and Ve maps, referring to the underlay information from the postcontrast imaging and excluding nonenhancing areas (Fig. 1). The mean values of these parameters were used for analysis. All DICOM data for DTI were transferred to a commercial software package (Nordic ICE, Nordic Imaging Lab). The parametric maps for the ADC and fractional anisotropy (FA) were coregistered with postcontrast T1weighted images. As with the permeability maps, ROIs were drawn to contain all enhancing components of the tumor in each section of the ADC and FA maps, referring to the underlay information from the postcontrast imaging and excluding nonenhancing areas (Fig. 1). The mean values of these parameters were used for analysis. We evaluated intraobserver agreements by comparing initial measurements with repeated measurements by a same investigator using the same technique.
Statistical Analysis
Imaging features were correlated with MGMT methylation status using chi-square tests or Fisher’s exact tests. Interobserver agreement for each imaging feature was calculated using the kappa statistic. A kappa value of 0.81–1.0 indicated excellent agreement between the 2 observers; 0.61–0.80 indicated good agreement, 0.41–0.6 indicated moderate agreement, 0.21–0.4 indicated fair agreement, and 0–0.2 indicated only slight agreement.8 Intraobserver agreements for quantitative imaging parameters were evaluated with the intraclass correlation coefficient. Based on normality testing, Mann-Whitney U-tests with Bonferroni correction were used to evaluate the relationships between MGMT methylation status and quantitative imaging parameters. We performed receiver operating characteristic curve analysis to find the optimal cutoff value for the presence of MGMT methylation. All statistical analyses were performed using statistical software (SAS version 9.2, SAS Institute Inc.; and MedCalc version 9.3.6.0, MedCalc Software); p values < 0.05 were considered significant.
Results
Among 43 patients, 16 (37.2%) cases of glioblastoma had confirmed MGMT methylation, comparable to rates of MGMT methylation in other studies.2,15 Qualitative Imaging Analysis and MGMT Methylation Status
Interobserver agreement for imaging features was good to excellent, with kappa values of 0.622 for tumor margin, 0.630 for enhancement pattern, 0.946 for edema, 0.839 for cysts, 0.78 for necrosis, 0.952 for nonenhancing tumor regions, and 0.788 for T2 signal intensity. Table 2 summarizes the imaging features that were qualitatively assessed on conventional imaging between the 2 glioblastoma groups. The MGMT methylation status was not significantly associated with any imaging features.
Quantitative Imaging Analysis and MGMT Methylation Status
Intraobserver agreements for quantitative imaging
Fig. 1. On ADC (B), FA (C), and permeability maps (D), ROIs (arrows) were drawn in all slices encompassing the entire tumor volume in the solid enhancing portion of the tumor but not in the nonenhancing portions of the tumor with T2 hyperintensity (not shown), as seen on a postcontrast T1-weighted image (A).
J Neurosurg / Volume 121 / August 2014
369
S. S. Ahn et al. TABLE 2: Correlations between MGMT methylation status and imaging features No. of Patients (%) Parameter tumor margins well defined poorly defined enhancement pattern ring nodular mixed edema yes no cyst yes no necrosis yes no nonenhancing tumor yes no T2 signal intensity homogeneous heterogeneous
Methylated MGMT
Unmethylated MGMT
Total
14/17 (82.4) 3/17 (17.6)
15/26 (57.7) 11/26 (42.3)
29/43 (67.4) 14/43 (32.6)
p Value 0.092
0.313 6/17 (35.3) 3/17 (17.6) 8/17 (47.1)
11/26 (42.3) 1/26 (3.8) 14/26 (53.8)
17/43 (39.5) 4/43 (9.3) 22/43 (51.2)
12/17 (70.6) 5/17 (29.4)
17/26 (65.4) 9/26 (34.6)
29/43 (67.4) 14/43 (32.6)
0.722
1 5/17 (29.4) 12/17 (70.6)
7/26 (26.9) 19/26 (73.1)
12/43 (27.9) 31/43 (72.1)
12/17 (70.6) 5/17 (29.4)
23/26 (88.5) 3/26 (11.5)
35/43 (81.4) 8/43 (18.6)
8/17 (47.1) 9/17 (52.9)
17/26 (65.4) 9/26 (34.6)
25/43 (58.1) 18/43 (41.9)
1/17 (5.9) 16/17 (94.1)
2/26 (7.7) 24/26 (92.3)
3/43 (7) 40/43 (93)
0.23
0.234
1
parameters ranged from 0.785 to 0.916 (intraclass correlation coefficients). Ktrans values were significantly higher in the MGMT methylated group (median 0.091 vs 0.053 min-1, p = 0.018) (Figs. 2 and 3). Kep values trended higher in the MGMT methylated group (median 0.392 vs 0.325 min-1, p = 0.168); Ve, ADC, and FA were not different between the 2 groups (Table 3). According to the receiver operating characteristic curve analysis, the optimal cutoff value for the presence of MGMT methylation was a Ktrans value > 0.086 min-1 with an area under the curve of 0.756, a sensitivity of 56.3%, and a specificity of 85.2% (Fig. 4).
Discussion
Our results indicate that Ktrans can be used to predict MGMT methylation status. Glioblastoma with MGMT methylation shows a higher Ktrans value than glioblastoma without MGMT methylation. Several studies have demonstrated that permeability measurements with DCE-MRI may be helpful in discriminating the malignant potential of glioma by showing that the Ktrans of low-grade gliomas is lower than that of high-grade gliomas.3,12,17 High-grade tumors with a higher proportion of immature vessels from neoangiogenesis have increased
Fig. 2. Glioblastoma with methylated MGMT. The tumor mainly involves the right insula and shows heterogeneous enhancement (A). The mean ADC value was 1458 × 10-6 mm2/sec (B), and the mean FA value was 0.127 (C). The permeability map demonstrates a markedly increased Ktrans (0.275 min-1) (D).
370
J Neurosurg / Volume 121 / August 2014
Prediction of MGMT methylation in glioblastoma using MRI
Fig. 3. Glioblastoma with unmethylated MGMT. The tumor is in the right frontal lobe and shows peritumoral edema (A). The mean ADC value was 1274 × 10-6 mm2/sec (B), and the mean FA value was 0.223 (C). The permeability map demonstrates a minimal increase in Ktrans (0.04 min-1) (D).
endothelial permeability, which facilitates the transfer of contrast agents from plasma into the extracellular space. Therefore, one might expect that glioblastomas with MGMT methylation, which have been well documented to have a better prognosis, would demonstrate less aggressive features and hence a lower level of endothelial permeability than glioblastoma without MGMT methylation. Although we cannot adequately explain the reason, one explanation for these results could be that an increased Ktrans in glioblastomas with MGMT methylation signifies easier penetration of temozolomide and thus results in increased therapeutic success. One study found longer survival in cases with higher Ktrans when only high-grade gliomas were considered.13 The authors postulated that the positive relationship between Ktrans and survival is due to improved drug delivery to tumor tissues. Likewise, increased Ktrans in glioblastomas with MGMT methylation might represent improved passage of temozolomide via leaky blood vessels, resulting in a better prognosis. Pseudoprogression, a transient radiological increase in contrast enhancement after concurrent radiochemotherapy that is consequent to the breakdown of the BBB, has been reported to be more frequent in glioblastomas with MGMT methylation; the overall survival of patients with pseudoprogression was significantly higher than in those without pseudoprogression.1 Our results support the high prevalence of pseudoprogression in glioblastoma with MGMT methylation because the favorable treatment response might have contributed to the increased endothelial permeability to chemotherapeutic
agents in this tumor. Although temozolomide can cross the BBB, increased permeability is thought to impact intratumoral concentrations of this drug.20 In our study, the ADC and FA values from DTI were not significantly different between the MGMT methylated and unmethylated groups. This result is discordant with findings from previous studies. Pope et al., for example, reported a lower ADC value in the MGMT methylated group,15 while Moon et al. reported that high-grade gliomas with MGMT methylation demonstrated higher ADC and lower FA values than those without MGMT methylation.14 However, the latter study included both WHO Grade III and IV tumors with a small number of patients in each group. One explanation for such variable results could be that ADC and FA values can be affected by multiple factors. For example, ADC is lowered by high cellularity but is increased by edema and necrosis, which are common in glioblastoma. Moreover, FA reflecting destruction of the white matter due to tumor infiltration shows potentially dramatic regional and anatomy-specific variations. In addition, these studies used variable quantitative imaging analysis methods. We used the mean ADC value from ROIs corresponding to the entirety of the enhancing tumor, while Pope at al. used mean values for the lower peak of the ADC fitted with a binormal distribution from ROIs corresponding to the entirety of enhancing tumor, and Moon et al. used minimum ADC values and ADC ratios from 6 ROIs. Further investigations are warranted to address the potential role of DTI in predicting MGMT methylation status.
TABLE 3: Differences in permeability and DTI parameters according to MGMT promoter methylation status Median (interquartile range) Parameter Ktrans (min ) Kep (min−1) Ve ADC (10−6 mm2/sec) FA (10−3) −1
Methylated MGMT (n = 16) 0.091 (0.056–0.108) 0.392 (0.267–0.408) 0.249 (0.163–0.371) 1272.4 (1041.6–1428.9) 186.1 (145.3–221.6)
Unmethylated MGMT (n = 27) 0.053 (0.036–0.083) 0.325 (0.267–0.408) 0.218 (0.135–0.28) 1162.2 (1037.9–1349) 161.8 (133–198.5)
p Value* 0.018 0.168 0.858 0.924 0.416
* Corrected p value for multiple comparisons.
J Neurosurg / Volume 121 / August 2014
371
S. S. Ahn et al. obtained mean values from ROIs containing all enhancing components of the tumor in each section rather than ROIs from representative sections, which might have mitigated some of the measurement error in the spatially heterogeneous glioblastoma.
Conclusions
Transfer constant (Ktrans) of DCE-MRI may serve as a potential imaging biomarker to predict MGMT methylation status preoperatively in glioblastoma. However, further investigation with a larger cohort is necessary. Disclosure
Fig. 4. Receiver operating characteristic curve analysis for the prediction of MGMT promoter methylation status. The optimal cutoff value for the presence of MGMT methylation was a Ktrans > 0.086 min-1 with an area under the curve of 0.756.
Previously, nodular and mixed nodular enhancement patterns and ill-defined enhancing tumor margins have been reported to be associated with MGMT methylation.5,6,14 However, none of our conventional imaging features were significantly associated with MGMT methylation status, with p values ranging from 0.092 to 1. In accordance with our results, Gupta et al.9 also reported that there were no significant differences in imaging features between the groups after qualitative assessment of conventional MRI. In practice, most suspicious tumors are biopsied, and imaging prediction of MGMT methylation may have limited clinical benefit. However, a recent study reported that MGMT expression and promoter methylation might vary throughout a single glioblastoma, and consequently, results might depend on the site of surgical sampling.4 Therefore, imaging biomarkers could be the only tool capable of resolving this intratumoral heterogeneity and might serve as a surrogate for histopathology when pathology sampling is suboptimal. Furthermore, it could be of value if future preoperative treatment regimens are developed. There are several issues with our study that need to be addressed. First, 21 patients underwent subtotal resection. However, a majority of the enhancing portions were removed, and we excluded 1 patient who underwent stereotactic biopsy, which assumes the histological specimen is representative of the entire tumor without a significant sampling error. Second, we analyzed quantitative parameters from areas with tumor enhancement only because this would be more reproducible than including areas of nonenhancing tumor, which can be ill defined or obscured by edema. Third, we did not measure interobserver variability for quantitative analysis. However, we 372
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Lee. Acquisition of data: Ahn, Chang, EH Kim. Analysis and interpretation of data: Ahn, Shin, SH Kim, DW Kim. Drafting the article: Ahn. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Lee. Statistical analysis: DW Kim. Study supervision: Lee. References 1. Brandes AA, Franceschi E, Tosoni A, Blatt V, Pession A, Tallini G, et al: MGMT promoter methylation status can predict the incidence and outcome of pseudoprogression after concomitant radiochemotherapy in newly diagnosed glioblastoma patients. J Clin Oncol 26:2192–2197, 2008 2. Carrillo JA, Lai A, Nghiemphu PL, Kim HJ, Phillips HS, Kharbanda S, et al: Relationship between tumor enhancement, edema, IDH1 mutational status, MGMT promoter methylation, and survival in glioblastoma. AJNR Am J Neuroradiol 33:1349–1355, 2012 3. Cha S, Yang L, Johnson G, Lai A, Chen MH, Tihan T, et al: Comparison of microvascular permeability measurements, K(trans), determined with conventional steady-state T1weighted and first-pass T2*-weighted MR imaging methods in gliomas and meningiomas. AJNR Am J Neuroradiol 27: 409–417, 2006 4. Della Puppa A, Persano L, Masi G, Rampazzo E, Sinigaglia A, Pistollato F, et al: MGMT expression and promoter methylation status may depend on the site of surgical sample collection within glioblastoma: a possible pitfall in stratification of patients? J Neurooncol 106:33–41, 2012 5. Drabycz S, Roldán G, de Robles P, Adler D, McIntyre JB, Magliocco AM, et al: An analysis of image texture, tumor location, and MGMT promoter methylation in glioblastoma using magnetic resonance imaging. Neuroimage 49:1398–1405, 2010 6. Eoli M, Menghi F, Bruzzone MG, De Simone T, Valletta L, Pollo B, et al: Methylation of O6-methylguanine DNA methyltransferase and loss of heterozygosity on 19q and/or 17p are overlapping features of secondary glioblastomas with prolonged survival. Clin Cancer Res 13:2606–2613, 2007 7. Esteller M, Herman JG: Generating mutations but providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human cancer. Oncogene 23:1–8, 2004 8. Fleiss JL, Levin B, Paik MC: The measurement of interrater agreement, in Statistical Methods for Rates and Proportions, ed 3. Hoboken, NJ: John Wiley & Sons, 2004, pp 598–626 9. Gupta A, Omuro AM, Shah AD, Graber JJ, Shi W, Zhang Z,
J Neurosurg / Volume 121 / August 2014
Prediction of MGMT methylation in glioblastoma using MRI et al: Continuing the search for MR imaging biomarkers for MGMT promoter methylation status: conventional and perfusion MRI revisited. Neuroradiology 54:641–643, 2012 10. Hegi ME, Diserens AC, Godard S, Dietrich PY, Regli L, Ostermann S, et al: Clinical trial substantiates the predictive value of O-6-methylguanine-DNA methyltransferase promoter methylation in glioblastoma patients treated with temozolomide. Clin Cancer Res 10:1871–1874, 2004 11. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M, et al: MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 352:997–1003, 2005 12. Jia Z, Geng D, Xie T, Zhang J, Liu Y: Quantitative analysis of neovascular permeability in glioma by dynamic contrastenhanced MR imaging. J Clin Neurosci 19:820–823, 2012 13. Mills SJ, Patankar TA, Haroon HA, Balériaux D, Swindell R, Jackson A: Do cerebral blood volume and contrast transfer coefficient predict prognosis in human glioma? AJNR Am J Neuroradiol 27:853–858, 2006 14. Moon WJ, Choi JW, Roh HG, Lim SD, Koh YC: Imaging parameters of high grade gliomas in relation to the MGMT promoter methylation status: the CT, diffusion tensor imaging, and perfusion MR imaging. Neuroradiology 54:555–563, 2012 15. Pope WB, Lai A, Mehta R, Kim HJ, Qiao J, Young JR, et al: Apparent diffusion coefficient histogram analysis stratifies progression-free survival in newly diagnosed bevacizumabtreated glioblastoma. AJNR Am J Neuroradiol 32:882–889, 2011 16. Riemenschneider MJ, Hegi ME, Reifenberger G: MGMT promoter methylation in malignant gliomas. Target Oncol 5: 161–165, 2010
J Neurosurg / Volume 121 / August 2014
17. Roberts HC, Roberts TP, Brasch RC, Dillon WP: Quantitative measurement of microvascular permeability in human brain tumors achieved using dynamic contrast-enhanced MR imaging: correlation with histologic grade. AJNR Am J Neuroradiol 21:891–899, 2000 18. Tofts PS, Kermode AG: Measurement of the blood-brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson Med 17:357–367, 1991 19. Weller M, Stupp R, Reifenberger G, Brandes AA, van den Bent MJ, Wick W, et al: MGMT promoter methylation in malignant gliomas: ready for personalized medicine? Nat Rev Neurol 6:39–51, 2010 20. Zhou Q, Guo P, Kruh GD, Vicini P, Wang X, Gallo JM: Predicting human tumor drug concentrations from a preclinical pharmacokinetic model of temozolomide brain disposition. Clin Cancer Res 13:4271–4279, 2007 Manuscript submitted October 14, 2013. Accepted May 5, 2014. This research was presented as a scientific poster at the Radio logical Society of North America’s 98th Scientific Assembly and Annual Meeting, November 27, 2012, Chicago, IL. Please include this information when citing this paper: published online June 20, 2014; DOI: 10.3171/2014.5.JNS132279. Address correspondence to: Seung-Koo Lee, M.D., Ph.D., Department of Radiology, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul 120-752, Republic of Korea. email: [email protected].
373
Prediction of methylguanine methyltransferase promoter methylation in glioblastoma using dynamic contrast-enhanced magnetic resonance and diffusion tensor imaging Clinical article Sung Soo Ahn, M.D., Ph.D.,1 Na-Young Shin, M.D.,1 Jong Hee Chang, M.D., Ph.D., 2 Se Hoon Kim, M.D., Ph.D., 3 Eui Hyun Kim, M.D., 2 Dong Wook Kim, Ph.D., 4 and Seung-Koo Lee, M.D., Ph.D.1 Department of Radiology, Research Institute of Radiological Science, Yonsei University College of Medicine; Departments of 2Neurosurgery and 3Pathology, Brain Research Institute, Yonsei University College of Medicine; and 4Biostatistics Collaboration Unit, Yonsei University College of Medicine, Seoul, Korea
1
Object. The methylation status of the methylguanine methyltransferase (MGMT) promoter has been associated with treatment response in glioblastoma. The authors aimed to assess whether MGMT methylation status can be predicted by dynamic contrast-enhanced (DCE) MRI and diffusion tensor imaging (DTI). Methods. This retrospective study included 43 patients with pathologically diagnosed glioblastoma who had undergone preoperative DCE-MRI and DTI and whose MGMT methylation status was available. The imaging features were qualitatively assessed using conventional MR images. Regions of interest analyses for DCE-MRI permeability parameters (transfer constant [Ktrans], rate transfer coefficient [Kep], and volume fraction of extravascular extracellular space [Ve]) and DTI parameters (apparent diffusion coefficient [ADC] and fractional anisotropy [FA]) were performed on the enhancing solid portion of the glioblastoma. Chi-square or Mann-Whitney tests were used to evaluate relationships between MGMT methylation and imaging parameters. The authors performed receiver operating characteristic curve analysis to find the optimal cutoff value for the presence of MGMT methylation. Results. MGMT methylation was not significantly associated with any imaging features on conventional MR images. Ktrans values were significantly higher in the MGMT methylated group (median 0.091 vs 0.053 min-1, p = 0.018). However, Kep, Ve, ADC, and FA were not significantly different between the 2 groups. The optimal cutoff value for the presence of MGMT methylation was Ktrans > 0.086 min-1 with an area under the curve of 0.756, a sensitivity of 56.3%, and a specificity of 85.2%. Conclusions. Ktrans may serve as a potential imaging biomarker to predict MGMT methylation status preoperatively in glioblastoma; however, further investigation with a larger cohort is necessary. (http://thejns.org/doi/abs/10.3171/2014.5.JNS132279)
Key Words • glioblastoma • O-6-methylguanine-DNA methyltransferase • magnetic resonance imaging • permeability • diffusion tensor imaging • oncology
M
ethylguanine methyltransferase (MGMT) is a key enzyme in DNA repair because it removes alkyl groups from the O6 position of guanine, an important site of DNA alkylation.7 Epigenetic silencing of the MGMT gene by promoter methylation is associated with low levels of MGMT proteins and consequently
Abbreviations used in this paper: ADC = apparent diffusion coefficient; BBB = blood-brain barrier; DCE = dynamic contrastenhanced; DTI = diffusion tensor imaging; FA = fractional anisotropy; Kep = rate transfer coefficient; Ktrans = transfer constant; MGMT = methylguanine methyltransferase; ROI = region of interest; Ve = volume fraction of extravascular extracellular space.
J Neurosurg / Volume 121 / August 2014
reduces DNA repair activity against alkylating agents.16,19 Therefore, glioblastoma with MGMT promoter methylation can be expected to have a more favorable response to temozolomide, a DNA alkylating agent. Several studies have reported that the methylation status of the MGMT gene promoter is associated with treatment response to chemotherapy and prognosis.10,11 MGMT methylation has been reported in 30%–60% of glioblastomas.19 Although genetic analysis using surgical specimens is the This article contains some figures that are displayed in color online but in black-and-white in the print edition.
367
S. S. Ahn et al. gold standard for assessing the methylation status of MGMT, there are limitations to this process. It requires an invasive procedure where there is still the possibility of incomplete biopsy sampling due to a spatially heterogeneous glioblastoma. Sufficient samples may be difficult to obtain if the tumor is inaccessible to the surgeon. Therefore, it would be helpful in predicting treatment response and prognosis if it were possible to noninvasively determine MGMT methylation status with preoperative imaging. A few studies have described distinctive radiological features of glioblastoma with MGMT methylation using conventional MRI and diffusion-weighted or diffusion tensor imaging (DTI).5,14,15 However, there are some discrepancies between the reports; some investigators have reported that the apparent diffusion coefficient (ADC) value is lower in the MGMT methylated group,15 whereas others have reported that the ADC ratio is significantly higher in the MGMT methylated group.14 While some studies have reported an enhancement pattern or tumor margin characteristic that appears to be associated with MGMT methylation status, others have not found such an association.5,6,9,14 Dynamic contrast-enhanced (DCE) MRI using a contrast agent is an emerging MRI technique based on kinetic modeling of microvascular permeability that enables quantification of blood-brain barrier (BBB) breakdown.18 Several studies have shown efficacy of permeability parameters from DCE-MRI in glioma grading.12,17 However, no study has evaluated the association between permeability parameters and MGMT methylation status in glioblastoma. Therefore, the aim of this study was to identify imaging biomarkers for MGMT methylation using quantitative parameters from DTI and DCE-MRI in addition to conventional imaging features.
Methods
The institutional review board approved this retrospective study and did not require patient approval or informed consent for the review of patient images. However, we did obtain informed consent for MGMT promoter gene evaluation at the time of surgery. Patients
We retrospectively studied 49 patients between October 2011 and January 2013 from the neurooncology database at our institution on the basis of the following criteria: 1) a histopathological diagnosis of glioblastoma based on the WHO grading system and 2) a record of preoperative MRI performed with DTI and DCE-MRI. We excluded 6 patients due to the following conditions: inadequate MRI (n = 3), failure to obtain informed consent for MGMT promoter gene evaluation (n = 2), or stereotactic biopsy (n = 1). Accordingly, 43 patients were enrolled (25 men and 18 women; mean age 58 ± 14.5 years). The characteristics of the 43 patients are shown in Table 1. The mean interval between preoperative MRI and surgery was 5.8 ± 4.3 days. Of the 43 patients, 18 underwent gross-total resection, 21 underwent subtotal resection, and 4 underwent partial resection. All pathological specimens were examined by an experienced neuropathologist (K.S.H.), and the MGMT methylation status
368
TABLE 1: Characteristics of 43 patients with glioblastoma* Parameter mean age sex M F extent of resection total subtotal partial
Methylated Unmethylated MGMT (n = 16) MGMT (n = 27)
Total (n = 43)
64.3 ± 9.6
54.2 ± 15.7
58 ± 14.5
9 (56.2) 7 (43.8)
16 (59.3) 11 (40.7)
25 (58.1) 18 (41.9)
8 (50) 7 (43.8) 1 (6.3)
10 (37) 14 (51.9) 3 (11.1)
18 (41.9) 21 (48.8) 4 (9.3)
* Data are presented as numbers of patients (%) unless noted otherwise. The mean value is presented as ± SD.
was assessed with methylation-specific polymerase chain reaction. Image Acquisition
Preoperative MR images were obtained using a 3.0-T system (Achieva, Philips) and an 8-channel SENSE head coil. The preoperative evaluation MRI protocol included the following conventional sequences: pre- and postcontrast T1-weighted imaging (TR 2000 msec, TE 10 msec, FOV 240 mm, slice thickness 5 mm, and matrix 256 × 256), T2-weighted imaging (TR 3000 msec, TE 80 msec, FOV 240 mm, slice thickness 5 mm, and matrix 256 × 256), and fluid-attenuated inversion recovery (TR 10,000 msec, TE 125 msec, FOV 240 mm, slice thickness 5 mm, and matrix 256 × 256). For DCE-MRI, precontrast 3D T1-weighted images were obtained with the following parameters: TR 6.3 msec, TE 3.1 msec, FOV 240 mm, matrix 192 × 192 mm, slice thickness 3 mm, and flip angle 5°. After the precontrast scan, 60 DCE T1-wighted images were obtained with the same MR parameters except for an increased flip angle of 15°. After acquisition of the fifth image volume, gadolinium-based contrast (0.1 ml/kg gadobutrol, Gadovist, Bayer Schering Pharma) was injected at a rate of 3 ml/sec. The acquisition time for DCE-MRI was 378 seconds. Diffusion tensor images were obtained by applying 6 different directions of orthogonal diffusion gradients and b values of 1000 seconds/mm2 and 0 seconds/mm2 (TR 8400 msec, TE 77 msec, FOV 240 mm, slice thickness 2 mm, and matrix 128 × 128). Qualitative Imaging Analysis
Two neuroradiologists (S.S.A. and N.Y.S.) who were blinded to the patients’ molecular and clinical data reviewed the conventional MR images on a standard picture archiving and communication system. All tumors were assessed for the following imaging features: enhancing tumor margin (well or poorly defined); enhancement patterns (ring, nodular, or mixed enhancement); presence of edema, cysts, necrosis, and nonenhancing tumor; and heterogeneity of the signal intensity on the T2-weighted images. Briefly, necrosis was defined as regions of peJ Neurosurg / Volume 121 / August 2014
Prediction of MGMT methylation in glioblastoma using MRI ripheral and irregular enhancement surrounding areas of T2 hyperintensity. Nonenhancing tumor regions were defined as areas of intermediate T2 signal intensity associated with a mass effect and architectural distortion. Discordant interpretations were resolved by consensus. Quantitative Imaging Analysis
An experienced neuroradiologist (S.S.A.) who was blinded to molecular and clinical data performed postprocessing. Permeability maps from DCE-MRI, including volume transfer constant (Ktrans), rate transfer coefficient (Kep), and volume fraction of extravascular extracellular space (Ve) maps, were generated by off-line Pride tools, based on the pharmacokinetic model of Tofts and Kermode,18 provided by Philips Medical Systems. Postprocessing comprised motion correction of pixels from dynamic images, T1 mapping using different flip angles (5° and 15°), registration of pixels on a T1 map, vascular input function estimation, and pharmacokinetic modeling. All of these processes were automatically performed by Pride tools except for the drawing of regions of interest (ROIs) for vascular input function. The ROIs for the vascular input function were drawn from the vertical part of the superior sagittal sinus on the middle section of the scanned volume. For volume-based analysis, ROIs were drawn to contain all enhancing components of the tumor in each section of the Ktrans, Kep, and Ve maps, referring to the underlay information from the postcontrast imaging and excluding nonenhancing areas (Fig. 1). The mean values of these parameters were used for analysis. All DICOM data for DTI were transferred to a commercial software package (Nordic ICE, Nordic Imaging Lab). The parametric maps for the ADC and fractional anisotropy (FA) were coregistered with postcontrast T1weighted images. As with the permeability maps, ROIs were drawn to contain all enhancing components of the tumor in each section of the ADC and FA maps, referring to the underlay information from the postcontrast imaging and excluding nonenhancing areas (Fig. 1). The mean values of these parameters were used for analysis. We evaluated intraobserver agreements by comparing initial measurements with repeated measurements by a same investigator using the same technique.
Statistical Analysis
Imaging features were correlated with MGMT methylation status using chi-square tests or Fisher’s exact tests. Interobserver agreement for each imaging feature was calculated using the kappa statistic. A kappa value of 0.81–1.0 indicated excellent agreement between the 2 observers; 0.61–0.80 indicated good agreement, 0.41–0.6 indicated moderate agreement, 0.21–0.4 indicated fair agreement, and 0–0.2 indicated only slight agreement.8 Intraobserver agreements for quantitative imaging parameters were evaluated with the intraclass correlation coefficient. Based on normality testing, Mann-Whitney U-tests with Bonferroni correction were used to evaluate the relationships between MGMT methylation status and quantitative imaging parameters. We performed receiver operating characteristic curve analysis to find the optimal cutoff value for the presence of MGMT methylation. All statistical analyses were performed using statistical software (SAS version 9.2, SAS Institute Inc.; and MedCalc version 9.3.6.0, MedCalc Software); p values < 0.05 were considered significant.
Results
Among 43 patients, 16 (37.2%) cases of glioblastoma had confirmed MGMT methylation, comparable to rates of MGMT methylation in other studies.2,15 Qualitative Imaging Analysis and MGMT Methylation Status
Interobserver agreement for imaging features was good to excellent, with kappa values of 0.622 for tumor margin, 0.630 for enhancement pattern, 0.946 for edema, 0.839 for cysts, 0.78 for necrosis, 0.952 for nonenhancing tumor regions, and 0.788 for T2 signal intensity. Table 2 summarizes the imaging features that were qualitatively assessed on conventional imaging between the 2 glioblastoma groups. The MGMT methylation status was not significantly associated with any imaging features.
Quantitative Imaging Analysis and MGMT Methylation Status
Intraobserver agreements for quantitative imaging
Fig. 1. On ADC (B), FA (C), and permeability maps (D), ROIs (arrows) were drawn in all slices encompassing the entire tumor volume in the solid enhancing portion of the tumor but not in the nonenhancing portions of the tumor with T2 hyperintensity (not shown), as seen on a postcontrast T1-weighted image (A).
J Neurosurg / Volume 121 / August 2014
369
S. S. Ahn et al. TABLE 2: Correlations between MGMT methylation status and imaging features No. of Patients (%) Parameter tumor margins well defined poorly defined enhancement pattern ring nodular mixed edema yes no cyst yes no necrosis yes no nonenhancing tumor yes no T2 signal intensity homogeneous heterogeneous
Methylated MGMT
Unmethylated MGMT
Total
14/17 (82.4) 3/17 (17.6)
15/26 (57.7) 11/26 (42.3)
29/43 (67.4) 14/43 (32.6)
p Value 0.092
0.313 6/17 (35.3) 3/17 (17.6) 8/17 (47.1)
11/26 (42.3) 1/26 (3.8) 14/26 (53.8)
17/43 (39.5) 4/43 (9.3) 22/43 (51.2)
12/17 (70.6) 5/17 (29.4)
17/26 (65.4) 9/26 (34.6)
29/43 (67.4) 14/43 (32.6)
0.722
1 5/17 (29.4) 12/17 (70.6)
7/26 (26.9) 19/26 (73.1)
12/43 (27.9) 31/43 (72.1)
12/17 (70.6) 5/17 (29.4)
23/26 (88.5) 3/26 (11.5)
35/43 (81.4) 8/43 (18.6)
8/17 (47.1) 9/17 (52.9)
17/26 (65.4) 9/26 (34.6)
25/43 (58.1) 18/43 (41.9)
1/17 (5.9) 16/17 (94.1)
2/26 (7.7) 24/26 (92.3)
3/43 (7) 40/43 (93)
0.23
0.234
1
parameters ranged from 0.785 to 0.916 (intraclass correlation coefficients). Ktrans values were significantly higher in the MGMT methylated group (median 0.091 vs 0.053 min-1, p = 0.018) (Figs. 2 and 3). Kep values trended higher in the MGMT methylated group (median 0.392 vs 0.325 min-1, p = 0.168); Ve, ADC, and FA were not different between the 2 groups (Table 3). According to the receiver operating characteristic curve analysis, the optimal cutoff value for the presence of MGMT methylation was a Ktrans value > 0.086 min-1 with an area under the curve of 0.756, a sensitivity of 56.3%, and a specificity of 85.2% (Fig. 4).
Discussion
Our results indicate that Ktrans can be used to predict MGMT methylation status. Glioblastoma with MGMT methylation shows a higher Ktrans value than glioblastoma without MGMT methylation. Several studies have demonstrated that permeability measurements with DCE-MRI may be helpful in discriminating the malignant potential of glioma by showing that the Ktrans of low-grade gliomas is lower than that of high-grade gliomas.3,12,17 High-grade tumors with a higher proportion of immature vessels from neoangiogenesis have increased
Fig. 2. Glioblastoma with methylated MGMT. The tumor mainly involves the right insula and shows heterogeneous enhancement (A). The mean ADC value was 1458 × 10-6 mm2/sec (B), and the mean FA value was 0.127 (C). The permeability map demonstrates a markedly increased Ktrans (0.275 min-1) (D).
370
J Neurosurg / Volume 121 / August 2014
Prediction of MGMT methylation in glioblastoma using MRI
Fig. 3. Glioblastoma with unmethylated MGMT. The tumor is in the right frontal lobe and shows peritumoral edema (A). The mean ADC value was 1274 × 10-6 mm2/sec (B), and the mean FA value was 0.223 (C). The permeability map demonstrates a minimal increase in Ktrans (0.04 min-1) (D).
endothelial permeability, which facilitates the transfer of contrast agents from plasma into the extracellular space. Therefore, one might expect that glioblastomas with MGMT methylation, which have been well documented to have a better prognosis, would demonstrate less aggressive features and hence a lower level of endothelial permeability than glioblastoma without MGMT methylation. Although we cannot adequately explain the reason, one explanation for these results could be that an increased Ktrans in glioblastomas with MGMT methylation signifies easier penetration of temozolomide and thus results in increased therapeutic success. One study found longer survival in cases with higher Ktrans when only high-grade gliomas were considered.13 The authors postulated that the positive relationship between Ktrans and survival is due to improved drug delivery to tumor tissues. Likewise, increased Ktrans in glioblastomas with MGMT methylation might represent improved passage of temozolomide via leaky blood vessels, resulting in a better prognosis. Pseudoprogression, a transient radiological increase in contrast enhancement after concurrent radiochemotherapy that is consequent to the breakdown of the BBB, has been reported to be more frequent in glioblastomas with MGMT methylation; the overall survival of patients with pseudoprogression was significantly higher than in those without pseudoprogression.1 Our results support the high prevalence of pseudoprogression in glioblastoma with MGMT methylation because the favorable treatment response might have contributed to the increased endothelial permeability to chemotherapeutic
agents in this tumor. Although temozolomide can cross the BBB, increased permeability is thought to impact intratumoral concentrations of this drug.20 In our study, the ADC and FA values from DTI were not significantly different between the MGMT methylated and unmethylated groups. This result is discordant with findings from previous studies. Pope et al., for example, reported a lower ADC value in the MGMT methylated group,15 while Moon et al. reported that high-grade gliomas with MGMT methylation demonstrated higher ADC and lower FA values than those without MGMT methylation.14 However, the latter study included both WHO Grade III and IV tumors with a small number of patients in each group. One explanation for such variable results could be that ADC and FA values can be affected by multiple factors. For example, ADC is lowered by high cellularity but is increased by edema and necrosis, which are common in glioblastoma. Moreover, FA reflecting destruction of the white matter due to tumor infiltration shows potentially dramatic regional and anatomy-specific variations. In addition, these studies used variable quantitative imaging analysis methods. We used the mean ADC value from ROIs corresponding to the entirety of the enhancing tumor, while Pope at al. used mean values for the lower peak of the ADC fitted with a binormal distribution from ROIs corresponding to the entirety of enhancing tumor, and Moon et al. used minimum ADC values and ADC ratios from 6 ROIs. Further investigations are warranted to address the potential role of DTI in predicting MGMT methylation status.
TABLE 3: Differences in permeability and DTI parameters according to MGMT promoter methylation status Median (interquartile range) Parameter Ktrans (min ) Kep (min−1) Ve ADC (10−6 mm2/sec) FA (10−3) −1
Methylated MGMT (n = 16) 0.091 (0.056–0.108) 0.392 (0.267–0.408) 0.249 (0.163–0.371) 1272.4 (1041.6–1428.9) 186.1 (145.3–221.6)
Unmethylated MGMT (n = 27) 0.053 (0.036–0.083) 0.325 (0.267–0.408) 0.218 (0.135–0.28) 1162.2 (1037.9–1349) 161.8 (133–198.5)
p Value* 0.018 0.168 0.858 0.924 0.416
* Corrected p value for multiple comparisons.
J Neurosurg / Volume 121 / August 2014
371
S. S. Ahn et al. obtained mean values from ROIs containing all enhancing components of the tumor in each section rather than ROIs from representative sections, which might have mitigated some of the measurement error in the spatially heterogeneous glioblastoma.
Conclusions
Transfer constant (Ktrans) of DCE-MRI may serve as a potential imaging biomarker to predict MGMT methylation status preoperatively in glioblastoma. However, further investigation with a larger cohort is necessary. Disclosure
Fig. 4. Receiver operating characteristic curve analysis for the prediction of MGMT promoter methylation status. The optimal cutoff value for the presence of MGMT methylation was a Ktrans > 0.086 min-1 with an area under the curve of 0.756.
Previously, nodular and mixed nodular enhancement patterns and ill-defined enhancing tumor margins have been reported to be associated with MGMT methylation.5,6,14 However, none of our conventional imaging features were significantly associated with MGMT methylation status, with p values ranging from 0.092 to 1. In accordance with our results, Gupta et al.9 also reported that there were no significant differences in imaging features between the groups after qualitative assessment of conventional MRI. In practice, most suspicious tumors are biopsied, and imaging prediction of MGMT methylation may have limited clinical benefit. However, a recent study reported that MGMT expression and promoter methylation might vary throughout a single glioblastoma, and consequently, results might depend on the site of surgical sampling.4 Therefore, imaging biomarkers could be the only tool capable of resolving this intratumoral heterogeneity and might serve as a surrogate for histopathology when pathology sampling is suboptimal. Furthermore, it could be of value if future preoperative treatment regimens are developed. There are several issues with our study that need to be addressed. First, 21 patients underwent subtotal resection. However, a majority of the enhancing portions were removed, and we excluded 1 patient who underwent stereotactic biopsy, which assumes the histological specimen is representative of the entire tumor without a significant sampling error. Second, we analyzed quantitative parameters from areas with tumor enhancement only because this would be more reproducible than including areas of nonenhancing tumor, which can be ill defined or obscured by edema. Third, we did not measure interobserver variability for quantitative analysis. However, we 372
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Lee. Acquisition of data: Ahn, Chang, EH Kim. Analysis and interpretation of data: Ahn, Shin, SH Kim, DW Kim. Drafting the article: Ahn. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Lee. Statistical analysis: DW Kim. Study supervision: Lee. References 1. Brandes AA, Franceschi E, Tosoni A, Blatt V, Pession A, Tallini G, et al: MGMT promoter methylation status can predict the incidence and outcome of pseudoprogression after concomitant radiochemotherapy in newly diagnosed glioblastoma patients. J Clin Oncol 26:2192–2197, 2008 2. Carrillo JA, Lai A, Nghiemphu PL, Kim HJ, Phillips HS, Kharbanda S, et al: Relationship between tumor enhancement, edema, IDH1 mutational status, MGMT promoter methylation, and survival in glioblastoma. AJNR Am J Neuroradiol 33:1349–1355, 2012 3. Cha S, Yang L, Johnson G, Lai A, Chen MH, Tihan T, et al: Comparison of microvascular permeability measurements, K(trans), determined with conventional steady-state T1weighted and first-pass T2*-weighted MR imaging methods in gliomas and meningiomas. AJNR Am J Neuroradiol 27: 409–417, 2006 4. Della Puppa A, Persano L, Masi G, Rampazzo E, Sinigaglia A, Pistollato F, et al: MGMT expression and promoter methylation status may depend on the site of surgical sample collection within glioblastoma: a possible pitfall in stratification of patients? J Neurooncol 106:33–41, 2012 5. Drabycz S, Roldán G, de Robles P, Adler D, McIntyre JB, Magliocco AM, et al: An analysis of image texture, tumor location, and MGMT promoter methylation in glioblastoma using magnetic resonance imaging. Neuroimage 49:1398–1405, 2010 6. Eoli M, Menghi F, Bruzzone MG, De Simone T, Valletta L, Pollo B, et al: Methylation of O6-methylguanine DNA methyltransferase and loss of heterozygosity on 19q and/or 17p are overlapping features of secondary glioblastomas with prolonged survival. Clin Cancer Res 13:2606–2613, 2007 7. Esteller M, Herman JG: Generating mutations but providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human cancer. Oncogene 23:1–8, 2004 8. Fleiss JL, Levin B, Paik MC: The measurement of interrater agreement, in Statistical Methods for Rates and Proportions, ed 3. Hoboken, NJ: John Wiley & Sons, 2004, pp 598–626 9. Gupta A, Omuro AM, Shah AD, Graber JJ, Shi W, Zhang Z,
J Neurosurg / Volume 121 / August 2014
Prediction of MGMT methylation in glioblastoma using MRI et al: Continuing the search for MR imaging biomarkers for MGMT promoter methylation status: conventional and perfusion MRI revisited. Neuroradiology 54:641–643, 2012 10. Hegi ME, Diserens AC, Godard S, Dietrich PY, Regli L, Ostermann S, et al: Clinical trial substantiates the predictive value of O-6-methylguanine-DNA methyltransferase promoter methylation in glioblastoma patients treated with temozolomide. Clin Cancer Res 10:1871–1874, 2004 11. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M, et al: MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 352:997–1003, 2005 12. Jia Z, Geng D, Xie T, Zhang J, Liu Y: Quantitative analysis of neovascular permeability in glioma by dynamic contrastenhanced MR imaging. J Clin Neurosci 19:820–823, 2012 13. Mills SJ, Patankar TA, Haroon HA, Balériaux D, Swindell R, Jackson A: Do cerebral blood volume and contrast transfer coefficient predict prognosis in human glioma? AJNR Am J Neuroradiol 27:853–858, 2006 14. Moon WJ, Choi JW, Roh HG, Lim SD, Koh YC: Imaging parameters of high grade gliomas in relation to the MGMT promoter methylation status: the CT, diffusion tensor imaging, and perfusion MR imaging. Neuroradiology 54:555–563, 2012 15. Pope WB, Lai A, Mehta R, Kim HJ, Qiao J, Young JR, et al: Apparent diffusion coefficient histogram analysis stratifies progression-free survival in newly diagnosed bevacizumabtreated glioblastoma. AJNR Am J Neuroradiol 32:882–889, 2011 16. Riemenschneider MJ, Hegi ME, Reifenberger G: MGMT promoter methylation in malignant gliomas. Target Oncol 5: 161–165, 2010
J Neurosurg / Volume 121 / August 2014
17. Roberts HC, Roberts TP, Brasch RC, Dillon WP: Quantitative measurement of microvascular permeability in human brain tumors achieved using dynamic contrast-enhanced MR imaging: correlation with histologic grade. AJNR Am J Neuroradiol 21:891–899, 2000 18. Tofts PS, Kermode AG: Measurement of the blood-brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson Med 17:357–367, 1991 19. Weller M, Stupp R, Reifenberger G, Brandes AA, van den Bent MJ, Wick W, et al: MGMT promoter methylation in malignant gliomas: ready for personalized medicine? Nat Rev Neurol 6:39–51, 2010 20. Zhou Q, Guo P, Kruh GD, Vicini P, Wang X, Gallo JM: Predicting human tumor drug concentrations from a preclinical pharmacokinetic model of temozolomide brain disposition. Clin Cancer Res 13:4271–4279, 2007 Manuscript submitted October 14, 2013. Accepted May 5, 2014. This research was presented as a scientific poster at the Radio logical Society of North America’s 98th Scientific Assembly and Annual Meeting, November 27, 2012, Chicago, IL. Please include this information when citing this paper: published online June 20, 2014; DOI: 10.3171/2014.5.JNS132279. Address correspondence to: Seung-Koo Lee, M.D., Ph.D., Department of Radiology, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul 120-752, Republic of Korea. email: [email protected].
373
