The aim of the current study was to catalog genomic and epigenomic abnormalities in newly diagnosed glioblastoma patients and determine the correlation among clinical, genetic, and epigenetic profiles and clinical outcome.
The aim of the current study was to catalog genomic and epigenomic abnormalities in newly diagnosed glioblastoma patients and determine the correlation among clinical, genetic, and epigenetic profiles and clinical outcome.
This study retrospectively included 68 consecutive patients who underwent surgical treatment and received standard radiotherapy with temozolomide (TMZ)-based chemotherapy. Of a total of 68 patients, 39 patients (57.4%) received interferon (IFN)-β in combination of TMZ.
The genetic and epigenetic alterations frequently observed were EGFR amplification (51.5%), TP53 mutation (33.8%), CDKN2A loss (32.4%), TP53 loss (16.2%), methylation of the MGMT promoter (33.8%) and IDH1 mutation (5.9%). Multivariate analysis revealed that methylated MGMT promoter and the combination of TMZ and IFN-β were independent prognostic factors associated with survival. The median survival time (MST) of the patients who received the combination of IFN-β and TMZ was significantly greater with 19.9 months as compared to the TMZ alone group (12.7 months). Notably, in even patients whose tumors had unmethylated MGMT promoter, the MST prolonged to 17.2 months when receiving TMZ with IFN-β, compared to 12.5 months in those receiving TMZ without IFN-β.
Taken together, addition of IFN-β for newly diagnosed primary GBM achieved a favorable outcome, particularly in patients with unmethylated MGMT promoter. Cancer 2011. © 2010 American Cancer Society.
Glioblastoma multiforme (GBM) is one of the most frequent primary brain tumors in the central nervous system in adults and is highly malignant, with a median survival time of about one year from diagnosis. This is despite aggressive treatment, surgery, postoperative radiotherapy, and adjuvant chemotherapy. An international randomized trial by the European Organization for Research and Treatment of Cancer/National Cancer Institute of Canada (EORTC/NCIC) comparing radiotherapy alone and concomitant radiotherapy and temozolomide (TMZ) clearly attested the benefits of adjuvant TMZ chemotherapy for GBM patients.1 Since then, TMZ has been the current first-line chemotherapeutic agent for GBM.
A subanalysis in this trial showed the effectiveness of epigenetic silencing of the MGMT gene by promoter methylation with longer survival in patients with primary GBM; it also suggested the benefits of combining chemotherapy using TMZ with radiotherapy.2
Furthermore, there have been recent attempts to comprehensively profile GBM genes by The Cancer Genome Atlas (TCGA) project and other groups.3, 4 Some genetic aberrations in GBM, such as TP53 mutation or deletion, NF1 deletion or mutation, and ERBB2 mutation, have been found to be more common than previously reported. In addition, novel molecular markers, such as frequent mutations of the IDH1 and IDH2 genes in secondary GBM have been discovered.5-7 These findings on mutations, genomic and epigenomic aberrations, and transcriptomal features in GBM might aid in understanding the classification of GBM and its further potential clinical implications.
However, the TCGA project included GBM patients who received surgical treatment, and detailed information on adjuvant chemoradiotherapy was not provided. Therefore, the close relationship between the gene profile provided by TCGA and chemotherapy regimens remains unknown.3
In this current study, we aimed to determine the correlation between clinical, genetic, and epigenetic profiles, and clinical outcome in newly diagnosed GBM patients who received TMZ-based chemotherapy. Interestingly, we found a significant beneficial outcome in patients receiving TMZ in addition to IFN-β. Moreover, our study discovered that GBM patients with the ummethylated O6-methylguanine-DNA methyltransferase (MGMT) promoter, in particular, showed benefits from IFN-β.
We retrospectively reviewed 68 consecutive patients with newly diagnosed primary GBM who underwent surgical treatment at several academic tertiary-care neurosurgical institutions: Nagoya University Hospital, Hamamatsu University Hospital, Oita University Hospital, and Shizuoka Cancer Center from May 2006 through June 2010 after TMZ was approved as the treatment agent for malignant gliomas by the National Ministry of Health and Welfare of Japan. The diagnosis of GBM was established by histological confirmation according to the WHO guidelines8, 9 independently by at least two expert neuropathologists. The clinical, operative, and hospital course records were reviewed. Information collected from clinical notes included patient demographics, pre- and postoperative neuroimaging, and adjuvant therapy. Preoperative Eastern Cooperative Oncology Group performance status (ECOG PS) scores were assigned by the clinician at the time of evaluation and were available in the chart for review for all patients. The study was approved by the institutional review board at each participating hospital and complied with all provisions of the Declaration of Helsinki.
After undergoing surgery, the patients received focal external-beam radiotherapy by conventional radiation planning to approximately 60 Gray (Gy) (±5% total dose), with daily concurrent TMZ at 75 mg/m2 throughout the course of radiotherapy.
All patients received the standard Stupp regimen.1 In the absence of grade 3 or 4 hematological excessive toxicity, TMZ administration was continued until clinical or radiological evidence of disease progression was observed. Of these 68 patients, 39 patients (57.4%) received adjuvant IFN-β treatment (Table 1). Patients in Nagoya University and Oita University received chemotherapy consisting of IFN-β. There were no significant differences in any of the clinical parameters and genetic, epigenetic parameters (i.e., age, sex, preoperative PS, tumor location, extent of resection, genetic and epigenetic alterations between the institutions using regimen with and without IFN-β. The IFN-β chemotherapy regimen comprised 3 million international units (MIU)/body administered intravenously on alternate days during radiotherapy and TMZ-induction chemotherapy.10, 11 At the end of the induction period, after a 4-week interval, the patients were administered 3 MIU/body of IFN-β on the first morning every 4 weeks during TMZ maintenance chemotherapy. In the case of tumor progression, salvage or second-line therapy was administered at the investigators' discretion; most patients received additional chemotherapy.
|Parameter||No. of Patients||%|
|Preoperative ECOG performance status|
|Preoperative ECOG performance status|
Both radiological and clinical findings were used to evaluate the response. Follow-up magnetic resonance imaging (MRI) was performed for alternate cycles. If the MRI showed continued increase in enhancement, the case was considered as tumor progression. If re-resection was performed for a recurrent mass lesion, histological interpretation formed the basis for definitive diagnosis (treatment-related necrosis vs recurrent tumor).
All patients provided their written informed consent for molecular studies of their tumor, and the protocol was approved by the ethics committee at each center. Sixty-eight brain tumor specimens were obtained at the time of first surgical resection.
Tumor tissue samples were immediately frozen and stored at −80°C until the extraction of genomic DNA. DNA was prepared using the QIAmp DNA Mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Placental DNA was used as the normal control. The amount of DNA obtained from the tumor was sufficient for the subsequent genomic and epigenomic analyses.
Multiplex ligation-dependent probe amplification (MLPA) was used for the determination of allelic losses and gains of the gene in the tumor samples. The analysis was performed using the SALSA MLPA KIT P088-B1 and P105-C1 in accordance with the manufacturer's protocol (MRC Holland, Amsterdam, Netherland).12-15 Information regarding the probe sequences and ligation sites can be found at www.mlpa.com. Amplification products were separated on an ABI® 3130 × I Genetic Analyzer (Applied Biosystems, Foster City, CA) and quantified with Genemapper 4.0 software (Applied Biosystems). Duplicate experiments were performed to obtain accurate MLPA values. Data analysis was performed with an original Excel-based program based on MRC-Holland's procedures. Normalization for sample data was first performed on control probes, and each tumor sample was then normalized using the data on 2 control samples, using peripheral blood DNA. Single regression for control and tumor data slope correction was performed. Abnormal/normal ratio limits were set at 0.65 and 1.3. Statistical analysis was performed using the same Coffalyser software.
Tumor DNA was modified with bisulfate using the EpiTect bisulfite kit (Qiagen, Courtaboeuf Cedex, France). Pyrosequencing technology was used to determine the methylation status of the CpG island region of MGMT as described previously.16, 17 We used the touchdown PCR method. The primer sequences used were the MGMT forward primer, 5′-TTGGTAAATTAAGGTATAGAGTTTT-3′, and the MGMT biotinylated reverse primer, 5′-AAA CAATCTACGCATCCT-3′. PCR included a denaturation step at 95°C for 30 s, followed by annealing at various temperatures for 45 s, and extension at 72°C for 45 s. After PCR, the biotinylated PCR product was purified as recommended by the manufacturer. In brief, the PCR product was bound to Streptavidin Sepharose HP (Amersham Biosciences, Uppsala, Sweden), and the Sepharose beads containing the immobilized PCR product were purified, washed, and denatured using 0.2 N NaOH solution and washed again. Next, 0.3 mM pyrosequencing primer was annealed to the purified single-stranded PCR product, and pyrosequencing was performed using the PSQ HS 96 Pyrosequencing System (Pyrosequencing, Westborough, MA). The pyrosequencing primer was 5′-GGAAGTTGGGAAGG-3′. Methylation quantification was performed using the provided software.
For IDH sequencing, a fragment 129 bp in length, spanning the sequence encoding the catalytic domain of IDH1, including codon 132, and a fragment 150 bp in length spanning the sequence encoding the catalytic domain of IDH2, including codon 172, were amplified. We applied touchdown PCR, using the standard buffer conditions: it comprised 5 ng of DNA and AmpliTaq Gold DNA Polymerase (Applied Biosystems) run for 16 cycles with denaturation at 95°C for 30 s, annealing at 65 to 57°C (decreasing by 0.5°C per cycle) for 30 s, and extension at 72°C for 60 s in a total volume of 12.5 μl and add 30 cycles with denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 60 s, ending with at 72°C for 7 min to complete extension.
Direct sequencing was performed using BigDye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems). The reactions were carried out using an ABI 3100 Genetic Analyzer (Applied Biosystems).
Statistical analysis was performed using the statistical software SPSS for Windows, version 17.0 (SPSS Inc, Chicago, Ill). The Mann-Whitney U test, χ2 test, and Fisher exact test were used to test for association of clinical variables and molecular markers. Survival was estimated by using the Kaplan-Meier method, and survival curves were compared by using the log-rank test. Progression-free survival (PFS) was calculated from the day of first surgery until tumor progression, death, or end of follow up. Overall survival (OS) was calculated from the day of first surgery until death or the end of follow up. Univariate and multivariate analyses were performed to test the potential influence of baseline characteristics on survival. The effect of each single molecular marker on PFS and OS was investigated using the Cox proportional hazards model, adjusting for the major clinical prognostic factors, including age at diagnosis (<40 vs ≥40, <60 vs ≥60 years), ECOG performance status score (ECOG PS; ≤1 vs >1), extent of resection (macroscopic [gross] total resection [GTR] vs non-GTR), tumor location (superficial vs deep), MGMT promoter methylation status, chromosome 1p loss of heterozygosity (LOH), 19qLOH, PTEN loss, CDKN2A loss, TP53 loss and mutation, ERBB2 amplification, EGFR amplification, IDH1 and IDH2 mutation, and adjuvant therapy (with IFN-β vs without IFN-β). Factors with no significant association with survival, at a level of more than 0.05 in the multivariate analysis, were eliminated. The remaining factors in the multivariate proportional hazard model (P <.05) were considered to be independent predictors of survival.
To assess for the treatment effects of TMZ with IFN-β versus TMZ without IFN-β for overall survival (OS), the hazard ratio was computed using a proportional hazard model by baseline characteristics in stratified analysis.
Between May 2006 and June 2010, 68 consecutive patients newly diagnosed with primary GBM were registered in this study. Their clinical characteristics are summarized in Table 1. This study group comprised 41 men and 27 women aged 12-84 years (median, 55). The median preoperative ECOG PS score at diagnosis was 1 (range, 0-3); the preoperative ECOG PS score was <1 in 45 patients (66.2%). All tumors were located in the supratentorial region: 50 tumors were located in the superficial area (cortical or subcortical area), and 18 were located in deep anatomical structures such as the basal ganglia and corpus callosum. No tumor was noted in the optic nerve, olfactory nerve, and pituitary gland on pretreatment MRI. No tumor dissemination was detected by MRI. Surgical GTR was achieved in 24 patients (35.3%), and 44 patients underwent non-GTR (64.7%). None of the patients had concurrent active malignancy, and the baseline organ function before chemotherapy was as follows: absolute WBC ≥3000/mm3 or neutrophil count ≥1,500/mm3, platelet count ≥100,000/mm3, hemoglobin ≥8.0 g/dl, AST less than 2.5 × the upper limit of normal (ULN), total bilirubin 2 × ULN, and creatinine 2 × ULN, and electrocardiogram showing no serious arrhythmia and no serious ischemic heart disease. All patients received the standard Stupp regimen,1 and among these, 39 patients were received combination treatment with IFN-β, as described in the method section.
Of 68 cases, we could obtain sufficient genetic and epigenetic information in all cases. We used direct sequencing for TP53 and IDH1/2. We employed MLPA for the analysis of 1p/19q LOH, loss of TP53, PTEN and CDKN2A, and amplification of ERBB2 and EGFR. MLPA is a multiplex PCR method that detects abnormal copy numbers of up to 50 different genomic DNA sequences simultaneously. When comparing MLPA to FISH, MLPA not only has the advantage of being a multiplex technique but also one in which very small (50-70 nt) sequences are targeted, enabling MLPA to identify the frequent, single gene aberrations that are very small to be detected by FISH. Furthermore, for the detection of EGFR amplification, MLPA can examine exons 1-8, 13, 16, and 22, while previously reported real-time PCR covers only exons 2, 17, and 25. In our preliminary experiments, MLPA was found to be approximately 80% consistent with the real-time PCR method (data not shown). Notably, the methylation status of the MGMT promoter was analyzed by quantitative pyrosequencing technology. Although methylation-specific PCR analysis of MGMT promoter methylation is a widely applicable biomarker for the clinical setting, it is non quantitative and bears a risk of false-positive or false-negative results, especially when the DNA quality and/or quantity is low. Recent attempts to remedy some of these deficiencies have led to the development of an alternative sequence-based approach for methylation analysis, known as pyrosequencing. Pyrosequencing yields continuous methylation values ranging from 0-100%. Based on our comparisons with standard methylation-specific PCR and immunohistochemical study using the anti-MGMT antibody, we determined 14% as the threshold distinguishing unmethylation and methylation of the MGMT promoter in a given tumor.
As indicated in Figure 1 and Table 3, the alterations frequently observed were EGFR amplification (51.5%), TP53 mutation (33.8%), CDKN2A loss (32.4%), TP53 loss (16.2%), methylation of the MGMT promoter (33.8%), and IDH1 mutation (5.9%). These findings were consistent with those in previous reports.3, 9, 20, 21
|Parameter||No.||Months of OS||Log-rank test: P|
|TP53 mutation or loss|
The median follow-up time was 16.7 months (range, 3.4-46.7 months). The median PFS for all patients was 9.2 months (95% confidence interval [CI], 5.7-12.7). The median OS of all patients was 17.1 months (95% CI, 15.5-18.7) (Figure 2A). The log-rank tests demonstrated that tumor localization (P = .032), the MGMT methylation status (P = .029), and TP53 mutation or loss (P = .035) were associated with the OS of patients with GBM (Figure 2B-D). These findings were similar to univariate analysis, where deep location (P = .035), unmethylated MGMT promoter (P = .033) and TP53 mutation or loss (P = .038) were identified as candidate variables for poorer OS (Figure 2). In contrast, well-established prognostic factors such as age, ECOG PS, and the extent of tumor resection did not influence the outcome in this clinical setting. Next, we established multivariate survival models for OS. The model was designed to consider each of these factors without considering the interaction terms. The independent prognostic factors for OS were methylated MGMT promoter (P = .016).
We analyzed whether the use of IFN-β affected the survival of consecutive GBM patients treated with TMZ-based chemotherapy. Of the total 68 patients, 39 (57.4%) received IFN-β in combination of TMZ. Interestingly, the median OS of the combination group was significantly greater with 19.9 months (95% CI, 15.3-24.5) as compared to the TMZ alone group, which was 12.7 months (95% CI, 10.5 to 14.9) (Figure 3A). The 12-month-survival rate was 67.6% for the standard TMZ-treated cohort, whereas it was 83.6% for the combination group. The 24-month survival rates were 22.1% and 34.5%, respectively, for the 2 groups. The difference was statistically significant as determined by the log-rank test and univariate and multivariate analyses.
Next, we sought to determine the subpopulation that had benefited from the use of the IFN-β combination treatment. It is well known that patients with GBM containing the methylated MGMT promoter benefit from TMZ, but those with the unmethylated MGMT promoter show no such benefits.1, 2 Consistently, the median OS of 45 patients with the unmethylated MGMT status was significantly lesser than that of the patients with the methylated promoter (median OS = 15.1 months; 95% CI, 11.3-18.9). Notably, even in patients whose tumors had the unmethylated MGMT promoter, the median OS was prolonged to 17.2 months (95% CI, 13.9-20.6) when receiving TMZ with IFN-β as compared to the 12.5 months (95% CI, 11.3-13.7) in those receiving TMZ without IFN-β (P = .017) (Figure 3B).
Various associations of these clinical and molecular parameters were evaluated. A complete overview of the pairwise associations between these parameters and chemotherapy with or without IFN-β is provided in Figure 4. The relative hazards of OS between TMZ with or without IFN-β groups according to 6 baseline covariates, calculated by means of multivariate analysis, are shown. There were significant associations among patients under 40 years of age (P = .025), with ECOG PS ≤1 (P = .004), deep tumor location (P = .028), non-GTR (P = .048), and ummethylated MGMT status (P = .02) (Figure 4).
In this study, we analyzed the genomic abnormalities in 68 consecutive newly diagnosed patients with GBM who were treated with TMZ-based chemotherapy. We observed TP53 mutation (33.8%), TP53 loss (16.2%), EGFR amplification (51.5%), CDKN2A loss (32.4%), and methylation of the MGMT promoter (33.8%). Recent large-scale efforts to characterize the GBM genome have identified additional alterations in genes not previously implicated in glioma, such as ERBB2 and IDH1/IDH2 mutation in primary and secondary GBM, respectively, and a significant incidence of mutation and genomic loss of NF1.3, 4, 6 The TCGA study also noted TP53 mutations and losses in 35% of the cases, which is a surprisingly higher frequency than that reported previously.3, 20, 21 Furthermore, this study also revealed EGFR amplification (45%), CDKN2A loss (52.0%), and methylation of the MGMT promoter (20.9%). These results were consistent with our data. IDH1 mutations have recently been identified in gliomas, which are a strong predictor of a more favorable prognosis.6 Our study supported the finding that within the group of primary GBM, IDH1 mutations are rare and tend to define a prognostically favorable outcome.
The current study demonstrated that the methylated MGMT promoter and the combination of IFN-β and TMZ were independent prognostic indicators of GBM patients on multivariate analysis. Epigenetic silencing by the MGMT promoter methylation correlates with improved survival in glioma patients treated with TMZ.2, 22-25 The prognostic significance of MGMT promoter methylation has been shown in several clinical trials. In these studies, MGMT promoter methylation was an independent favorable prognostic factor and patients whose tumor contained a methylated MGMT promoter showed overall prolonged survival when treated with TMZ and radiotherapy. Our results demonstrated similarly that MGMT promoter hypermethylation determined by a novel pyrosequencing technology was significantly associated with better OS.
There are several contradicting reports on survival related to the prognostic value of TP53 mutations in GBM, showing either no association or that the presence of TP53 mutations was a favorable or an unfavorable prognostic factor.9, 20, 21, 26 On the other hand, our results demonstrated that TP53 mutation or loss was significantly associated with poor OS only in univariate analysis, but not in multivariate analysis. These findings were not in conflict with recent evidence, which shows that TP53 mutations not only disrupt its function but also possess gain-of-function and dominant-negative effects on the wild-type p53 protein, thus making the mutated TP53 gene an oncogene.27
The current study demonstrated that newly diagnosed primary GBM patients were associated with a favorable outcome on IFN-β and TMZ combination chemotherapy. The IFN-β and TMZ combination group achieved a median OS of 19.9 months (Figure 3A). This excellent result was almost equal to the median OS of only patients with the methylated MGMT promoter in the EORTC/NCIC trial.
IFN-β elicits pleiotropic biological effects such as antiproliferation, immunomodulation, and cell differentiation.28 Furthermore, it has been widely used either alone or in combination with other antitumor agents in the treatment of malignant brain tumors and melanomas. In our previous studies, we showed that combination therapy with IFN-β and nitrosourea has been particularly useful in the treatment of malignant gliomas in Japan.10 IFN-β has multifaceted functions related to antitumor activity, such as cytostatic effects, participating in the differentiation of CTLs and potentiation of their antitumor immunological responses, and behavior as a drug sensitizer to enhance toxicity against various malignant neoplasms when administered in combination with nitrosourea.10 Previously, in an in vitro study, we corroborated that IFN-β markedly enhanced chemosensitivity to TMZ29; this manifestation revealed that one of the major mechanisms by which IFN-β enhances chemosensitivity is the down-regulation of MGMT transcription. This effect was also confirmed in an experimental animal model.30 A subanalysis in this study showed that patients whose tumor had an unmethylated promoter benefited from the addition of IFN-β, suggesting that the combination of IFN-β and TMZ might provide better clinical outcomes in patients with the unmethylated MGMT promoter (Figures 3B, 4). Although we discovered that the patients under 40 years of age at diagnosis and those who had an initial ECOG PS ≤1 seemed to receive the benefit from IFN-β and TMZ combination therapy, our phase I study revealed that the combination regimen of IFN-β and TMZ was safe and well tolerated even in patients with older age and worse PS (Figure 4; manuscript in submission). In addition, the benefit associated with IFN-β was shown in patients whose tumors were deep, who had undergone non-GTR (Figure 4). This finding suggests that IFN-β might be better for use in cases of complicated tumor removal, i.e., when the tumors were deep, all the tumors could not be removed because they were, for example, located in an eloquent area or around essential structures.
In summary, this study supported the hypothesis that in cases of newly diagnosed primary GBM, IFN-β and TMZ combination therapy was significantly associated with a favorable outcome. To our knowledge, this is the first study to associate the survival benefits derived from IFN-β and TMZ combination. These benefits were, in particular, well correlated in patients with an unmethylated MGMT promoter.
Our results are limited as opposed to a prospective clinical trial as retrospective studies might have been influenced by unrecognized biases. However, the subject group we used was a consecutive series of patients, and this study provides novel information on the treatment for GBM. Thus, accumulation of evidence for this treatment will help further improvement of this disease and hopefully become a novel therapy. We are planning a prospective randomized control trial to compare the clinical outcomes between TMZ alone and a combination of TMZ and IFN-β in newly diagnosed GBM patients.
We thank Mr. Akiyoshi Sakai (Clinical Laboratory, Kariya Toyota General Hospital, Kariya, Japan), Mr. Hideaki Maruse, Mr. Takafumi Fukui, and Mr. Yosuke Furui (FALCO biosystems, Kyoto, Japan) for wonderful technical assistance.
This work was supported by grants from a Grant-in-Aid (C) for Scientific Research from the Ministry of Health, Labor, and Welfare, Japan (A.N.).