OBJECTIVE: Multidrug resistance (MDR) remains a major obstacle to successful chemotherapeutic treatment of cancer and can be caused by overexpression of P-glycoprotein, the MDR1 gene product. P-glycoprotein (P-gp) recognizes several chemotherapeutic agents as a substrate and prevents their accumulation. Among them, both technetium-99-sestamibi and technetioum-99-tetrofosmin ( 99m Tc-TF) are single photon emission computed tomography tracers. 99m Tc-MIBI uptake has been proven in vitro and in

more »

... o to inversely correlate with P-gp levels of tumors. In vitro studies have shown that 99m Tc-TF is influenced in a lesser degree from the P-gp expression, thus may be superior to 99m Tc-MIBI for brain tumor imaging. In the present study, we evaluated in vivo whether 99m Tc-TF uptake correlates with P-gp levels in gliomas. MATERIALS AND METHODS: Eighteen patients (10 males, 8 females, mean age 57.3 years) with histologically confirmed glioma were included in the study. There were 13 glioblastoma multiforme cases, 2 anaplastic astrocytomas, 2 anaplastic oligodendrogliomas, and 1 low grade astrocytoma. Brain SPECT by 99m Tc-TF was performed within a week prior to surgical excision and the expression of P-gp was assessed by immunohistochemistry. Radiotracer accumulation was assessed by a semiquantitative method of image analysis, calculating the lesion-to-normal (L/N) uptake ratio. RESULTS: The tracer uptake ranged from faint to profound (mean L/ N ¼ 8.2, range 1.8-20). The P-gp expression ranged from 0% to 45%. Using Spearman's rho analysis we found no correlation between tracer uptake (L/ N) and P-gp expression (P ¼ .21, r ¼ .538). CONCLUSION: The present data suggest that 99m Tc-TF uptake is not influenced by P-gp expression in gliomas. Thus, 99m Tc-TF constitute a suitable radiotracer for gliomas imaging. O.02. PERFUSION MR IN DIFFERENTIATING BETWEEN OBJECTIVES: The aim of this study was to evaluate perfusion magnetic resonance imaging (pMRI) for differentiation of tumor progression (PD) from pseudo-progression (Ps-PD) in patients with recurrent glioblastoma multiforme (GBM) following chemoradiation. BACKGROUND: The appearance of Ps-PD on brain MR following initial chemoradiotherapy is difficult to distinguish from true PD. We examined whether the technique of pMRI allows proper distinction between PD and Ps-PD in patients with recurrent GBM. METHODS: All files of patients with GBM with signs of radiological progression at T1-weighted gadolinium-enhanced MR scan were selected from the neuro-oncology clinic, followed by pMRI and 1 MR scan including pMRI 3 months later. Patients had received radiotherapy (60 Gy in 30 fractions) with concomitant (75 mg/ m 2 per day and adjuvant 200 mg/m 2 , Days 1 -5, q 4 wks) temozolomide. Clinical data and MR characteristics (localization, size, rCBV) were scored at radiological progression and 3 months later. Ps-PD was defined as absence of signs of PD at re-operation, no further progression, or spontaneous improvement MR 3 months later, and no new anti-tumor therapy or any increase dexamethasone dosing. MR findings including relative cerebral blood volume (rCBV) were analyzed and compared with clinical data. RESULTS: In 34 patients, 82% were diagnosed as PD and 18% as Ps-PD. In 32 of 34 patients, pMRI was evaluable. After establishing a cut-off value of 2.12 for rCBV, pMRI could differentiate between the two entities with a positive predictive value of 96% for the presence of true progression (PD). Sensitivity was 88% and specificity was 83% with a negative predictive value of 63% (P , .002, Fisher's exact test). CONCLUSION: pMRI seems to be a reliable technique to distinguish PD from Ps-PD in patients with recurrent GBM, and these results deserve further testing in larger sample for confirmation. O.03. ASSESSMENT OF BEVACIZUMAB/IRINOTECAN RESPONSE IN MALIGNANT GLIOMA BY ADC MAP IMAGE ANALYSIS BACKGROUND: Response assessment in malignant glioma following antiangiogenic treatment is challenging for conventional MR imaging (MRI). Despite decreased contrast-enhancement, non-enhancing parts of the tumor may continue to grow. In this retrospective study, we analyzed patients with recurrent malignant glioma during Bevacizumab/Irinotecan therapy using ADC map imaging analysis from diffusion-weighted MRI to yield ultrastructural information on cellular density and properties of the extracellular matrix in relation to the progression-free survival. METHODS: Fifteen patients treated with Bevacizumab/Irinotecan for recurrent malignant glioma were investigated by MRI every 2-3 months until tumor progression. Applying image segmentation, volumes of contrast-enhanced lesions on T1 and hyperintense nonenhancing T2 lesions were calculated. T2 hyperintense lesions were defined as regions of interest (ROIs) and registered to the corresponding ADC maps (T2-ADC). Histograms and cumulative histograms of the T2-ADC ROIs were calculated to quantify the apparent gray scale value distribution and were compared with progression-free survival. Software programs were used to perform segmentation (ITK-Snap), calculation of T2-ADC histograms (ImageJ), and statistical figures (SPSS). RESULTS: At 3-month follow-up, the overall mean contrast-enhanced T1 volume (in cm 3 ) decreased significantly from 26.8 (+29.43) (P ¼ .02) to 5.45 (+7.13). According to Mac Donald criteria, 12 patients showed response and 3 patients progressed. During the same period of time, the mean T2 volume (in cm 3 ) was significantly reduced in 8 cases (P ¼ .005) from 127.32 (+59.01) to 85.61 (+42.12) and increased in 7 cases (P ¼ .08) from 140.93 (+50.91) to 203.22 (+126.52). T2-ADC cumulative histograms showed differences in terms of gradient and kurtosis. In 8 cases an increasing gradient and high kurtosis represented an increased amount of low ADC grey scale values that can be interpreted as an augmentation of cellular density of the tumor. These patients showed a lower chance of progression-free survival compared with patients (n ¼ 6) with a decreasing slope and low kurtosis of the T2-ADC cumulative histograms. CONCLUSION: Changes in grey scale distribution in ADC cumulative histograms in patients with malignant recurrent glioma may be predictive for anti-angiogenic treatment response. BACKGROUND: Glioblastoma (GBM) is defined pathologically as an infiltrative glioma, and salvage therapy with bevacizumab is believed to increase the incidence of diffuse and distant invasion as assessed radiographically. PATIENTS AND METHODS: 80 adult patients with glioblastoma were treated with surgery followed by radiotherapy (RT) and concurrent and adjuvant temozolomide (TMZ). At first recurrence, 80 patients were treated with single agent, bevacizumab. At time of progression, 57 patients were treated with bevacizumab and a cytotoxic chemotherapy, cytotoxic chemotherapy alone, or on an investigational trial. Magnetic resonance imaging (MRI) were analyzed at four time points in each patient: at presentation, at first, second, and third recurrence. Four patterns of radiographic disease were assessed, local (unifocal disease), distant (second lesion noncontiguous with primary lesion), multifocal (.2 lesions including leptomeningeal dissemination), and diffuse. RESULTS: At presentation, 87.5% of glioblastoma were local, 6.25% distant, 3.75% multifocal, and 2.5% diffuse. At first recurrence following progression on RT/TMZ and before initiation of bevacizumab, 80% were local, 7.5% distant, 6.25% multifocal (including 1 with CSF dissemination), and 6.25% diffuse. At second recurrence following progression on bevacizumab, 71.25% were local, 8.75% distant, 8.75% multifocal (2 of 7 with CSF dissemination), and 11.25% were diffuse. At third recurrence (57 patients evaluable), 71.25% were local, 7.0% distant, 7.0% multifocal, and 14.0% were diffuse. Survival following progression on bevacizumab did not differ by pattern of radiographic recurrence. CONCLUSIONS: A majority of adult patients with GBM at diagnosis manifest MRI-defined local disease and maintain this pattern notwithstanding multiple recurrences and treatment with bevacizumab. OBJECTIVE: L-[Methyl-11 C]methionine (MET) positron emission tomography (PET), [ 18 F]-fluoro-3 ′ -deoxy-3 ′ -l-fluorothymidine (FLT) PET, and [ 18 F]-fluoromisonidazole (FMISO) are sensitive modalities for visualizing primary brain tumors. The objectives of this study were to evaluate the relationships between the uptake of MET, FLT, or FMISO and the histopathological grading in gliomas. METHODS: We examined 51 patients (22 males, 29 females; mean age: 48.7 years; range: 2-89 years; 10 diffuse astrocytomas, 1 oligodendroglioma, 3 oligoastrocytomas, 13 anaplastic astrocytomas, 1 anaplastic oligodendroglioma, 1 anaplastic ependymoma, and 22 glioblastomas) using MET-PET studies preoperatively. Moreover, 35 patients (17 males, 18 females; mean age: 51.4 years; range: 22-81 years; 6 diffuse astrocytomas, 1 oligodendroglioma, 7 anaplastic astrocytomas, and 21 glioblastomas) were examined with FLT-PET. Finally, 10 patients (6 males, 4 females; mean age: 55.8 years; range: 30-72 years; 1 diffuse astrocytomas, 2 anaplastic astrocytomas, 7 glioblastomas) were examined with FMISO-PET. MET, FLT, and FMISO uptakes were assessed by standardized uptake value of the tumor showing the maximum uptake (SUV max ), and the ratio of tumor tissue to the contralateral normal gray matter (T/N ratio). The tumor activity and degree of malignancy were evaluated using Ki-67 index. The correlations between SUV max and Ki-67 index were determined for all subjects. RESULTS: All glioblastomas showed tumor uptake of MET, FLT, and FMISO. The difference in MET T/N ratio was statistically significant between grades II and IV gliomas, but not significant between grades II and III gliomas. The difference in FLT T/N ratio was statistically significant between grades III and IV gliomas, but not significant between grades II and III gliomas. The difference in FMISO T/N ratio was statistically significant between grades III and IV gliomas. FLT SUV max in the tumor had a stronger correlation with Ki-67 index than MET SUV max . CONCLUSIONS: PET studies using MET, FLT, and FMISO are useful for preoperative diagnosis in gliomas. FLT-PET seems to be superior to MET-PET in assessment of the proliferation activity on gliomas of different grades. FMISO-PET is useful for non-invasive assessment of hypoxia in malignant gliomas. Advances in molecular imaging such as PET imaging techniques will facilitate more safe and solid management and therapy for the patients with malignant gliomas. INTRODUCTION: Hypermethylation of the MGMT promoter is strongly associated with longer progression-free and overall survival in glioblastoma (GBM) patients treated with radiotherapy and concomitant and adjuvant temozolomide. We here addressed the question whether GBM relapses are associated with changes in the MGMT promoter methylation status or altered expression of the DNA mismatch repair genes MLH1, MSH2, MSH6, or/and PMS2. METHODS: MGMT promoter methylation status was analyzed in paired primary and recurrent GBM samples of 80 patients using nonquantitative methylation-specific PCR (MSP). The vital tumor cell content of each primary and recurrent tumor specimen was histologically determined. Quantitative promoter methylation analyses using DNA pyrosequencing were performed, too, for MGMT in 48 patients as well as for the DNA mismatch repair genes MLH1, MSH2, MSH6, and PMS2 in 42 patients. Furthermore, the levels of MGMT, MLH1, MSH2, MSH6, and PMS2 proteins were analyzed semiquantitatively by immunohistochemistry (IHC) in 43 patients. RESULTS: MSP revealed MGMT promoter hypermethylation in 27 patients, borderline methylation in 3 patients, and no methylation in 50 patients at diagnosis. In 73 of the 80 patients, the MGMT promoter status of the primary tumor was retained at recurrence. In 6 patients, loss or reduction of MGMT promoter methylation was detected in the recurrent tumor; however, in 3 patients, this finding was explained by low tumor cell contents in the recurrent tumor specimens. In 1 patient, a change from MGMT borderline methylation (8% methylated alleles) to hypermethylation (50% methylated alleles) was detected. None of the investigated primary and recurrent glioblastomas showed MLH1, MSH2, MSH6, or PMS2 promoter hyperhypermethylation. However, immunohistochemical expression scores for MLH1, MSH2, MSH6, and PMS2 proteins were frequently reduced in the recurrent tumor when compared with the corresponding primary tumor. CONCLUSION: The MGMT promoter methylation status does not change from the primary to the recurrent tumor in the vast majority of GBM patients. Accordingly, changes in the MGMT promoter methylation status are unlikely to account for acquired resistance to temozolomide. Our results further suggest that GBM recurrences often demonstrate lower MLH1, MSH2, MSH6, and/ or PMS2 immunoreactivity scores. However, MLH1, MSH2, MSH6, and PMS2 promoter hypermethylation does not appear to account for these decreased protein levels and is appeared to be involved in not linked to GBM recurrence. Accordingly, the changes in the MGMT promoter methylation status are unlikely to account for acquired resistance to temozolomide. O.08. MGMT MRNA TRANSCRIPTIONAL ACTIVITY PREDICTS CLINICAL OUTCOME IN MALIGNANT GLIOMAS AFTER RADIO-/CHEMOTHERAPY OBJECTIVE: Epigenetic silencing of the gene that encodes for O 6 -methylguanine-DNA-methyltransferase (MGMT) has been correlated with favorable clinical outcome in patients with malignant gliomas after radio-/chemotherapy using alkylating agents such as temozolomide; it has been hypothesized that MGMT promoter methylation leads to a decreased MGMT mRNA transcriptional activity and subsequent protein expression with a diminished ability of the tumor to repair chemotherapy-induced lesions. The real clinical impact of this assumption, however, still remains controversial. In the current prospective study, we analyzed the predictive impact of MGMT mRNA expression in malignant glioma (64 patients) after radiotherapy and/or chemotherapy with temozolomide and its correlation with the MGMT promoter methylation status. METHODS: Only vital tumor samples harvested from open Abstracts iii2 NEURO-ONCOLOGY † S E P T E M B E R 2 0 1 0 Downloaded from https://academic.oup.com/neuro-oncology/article-abstract/12/suppl_3/iii1/1113158 by guest on