Virchows Arch DOI 10.1007/s00428-014-1622-6
ORIGINAL ARTICLE
MGMT promoter methylation and correlation with protein expression in primary central nervous system lymphoma L. Toffolatti & E. Scquizzato & S. Cavallin & F. Canal & M. Scarpa & P. M. Stefani & F. Gherlinzoni & A. P. Dei Tos
Received: 30 November 2013 / Revised: 4 June 2014 / Accepted: 30 June 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract The O6-methylguanine-DNA-methyltransferase (MGMT) gene encodes for a DNA repairing enzyme of which silencing by promoter methylation is involved in brain tumorigenesis. MGMT promoter methylation represents a favorable prognostic factor and has been associated with a better response to alkylating agents in glioma and systemic lymphoma. Primary central nervous system lymphoma (PCNSL) is a rare and aggressive extranodal malignant lymphoma. The current standard of care, based on high-dose methotrexate chemotherapy, has improved prognosis but outcome remains poor for a majority of patients. Therapeutic progress in this field is conditioned by limited biological and molecular knowledge about the disease. Temozolomide has recently emerged as an alternative option for PCNSL treatment. We aimed to analyze the MGMT gene methylation status in a series of 24 PCNSLs, to investigate the relationship between methylation status of the gene and immunohistochemical expression of MGMT protein and to evaluate the possible prognostic significance of these biomarkers. Our results confirm that methylation of the MGMT gene and loss of MGMT protein are frequent events in these lymphomas (54 % of our cases) and suggest that they are gender and age related. MGMT methylation showed high correlation with loss of protein expression (concordance correlation coefficient= −0.49; Fisher exact test: p60 months in three of four patients). The prognostic significance of these molecular markers in PCNSL needs to be further studied in groups of patients treated in a homogeneous way. Keywords Primary central nervous system lymphoma . MGMT . Promoter methylation . Immunohistochemistry . Alkylating agents
Introduction The O6-methylguanine-DNA-methyltransferase (MGMT) gene encodes for a DNA repairing enzyme that removes alkyl adducts from the O6 position of guanine in DNA. MGMT is ubiquitously expressed in normal human tissues [1] while its expression in tumor cells has been shown to be considerably heterogeneous. Loss of MGMT function in human cancer is mainly due to gene silencing by hypermethylation of CpG islands located in the promoter sequence [2]. It has been demonstrated that silencing of the MGMT gene is involved in lung, brain and colorectal tumorigenesis [3, 4]. MGMT promoter methylation status and MGMT protein expression have been extensively studied and characterized in glioblastoma, anaplastic glioma and low grade glioma. In these tumors, the methylation is not only associated with a better response to alkylating agents [5–8] but also represents a favorable prognostic factor irrespective of treatment [9, 8]. In further studies, MGMT gene has been found to be methylated in some B-cell lymphomas, including diffuse large B-cell lymphoma (DLBCL) and its methylation was associated with a better response to alkylating agents [10, 11]. The most commonly used method to determine epigenetic silencing of genes is methylation-specific PCR (MSP) that can
Virchows Arch
be applied to formalin-fixed and paraffin-embedded tissue but does not allow quantitative assessment of methylation [12]. Several alternative methods have been developed, giving a quantitative value of methylation [13, 14]. Loss of MGMT protein expression detected by immunohistochemistry (IHC) has been described in a number of studies, and has shown significantly discordant results with MGMT promoter methylation as well as with survival, not only in brain tumors but also in other neoplasms [15–17]. The unreliability of IHC has been explained by tumor heterogeneity, interobserver variability, lack of cut-off values and admixture with non-neoplastic MGMT expressing cells [18]. Primary central nervous system lymphomas (PCNSLs) are extranodal malignant lymphomas arising within the brain, eyes, leptomeninges, or spinal cord in the absence of systemic lymphoma at the time of diagnosis [19]. Currently, PCNSL are estimated to account for up to 1 % of non-Hodgkin lymphomas (NHLs) and 2–3 % of all primary brain tumors [20]. Although the outcome remains poor for the majority of patients, for a selected subgroup of patients the addition to whole-brain radiotherapy (WBRT) of high-dose methotrexate (MTX) chemotherapy, followed by autologous stem cell transplantation (ASCT), has considerably improved prognosis. In such patients a 3-fold longer median overall survival compared with WBRT alone is attained and this represents the current standard of care [21–23] However, this combined treatment exposes patients, especially the elderly, to a high risk for delayed neurotoxicity and some patients cannot receive the treatment because of impaired renal function or other comorbidity [23, 24]. Since temozolomide demonstrated good penetration of the blood–brain barrier, activity in a variety of brain tumors and favorable toxicity profile, it has recently emerged as an alternative option for PCNSL, as either first-line treatment in elderly or salvage therapy after failure to high-dose MTX [25–27]. Although the majority of patients did not achieve a complete response, these studies seem to identify a subgroup of patients that show an extended progression free survival after temozolomide treatment. Recently, MGMT promoter methylation status in PCNSLs has been analyzed in order to evaluate its possible role in predicting response to temozolomide treatment [27, 28]. Another study investigated the predictive value of MGMT expression by IHC [29]. Despite the small sample size, MGMT status seems to be related with response to temozolomide in PCNSLs. Our study is aimed to analyze the methylation status of the MGMT gene in a series of 24 primary CNS lymphomas applying a standardized quantitative real-time MSP assay (MethyLight test) and comparing the results with a standard protocol of non-quantitative MSP. The relationship between the methylation status of the MGMT gene and the immunohistochemical expression of MGMT protein was investigated, and the possible prognostic significance of these biomarkers was evaluated.
Materials and methods Case selection and pathological evaluation We retrospectively searched the pathology records of the Pathology Department at the Treviso General Hospital for patients with PCNSL who underwent stereotactic biopsy or surgery from 2000 to 2011. All the original slides of available cases were reviewed by one of the authors (F.C.) for confirmation of diagnoses and evaluation of quality and quantity of neoplastic tissue. Formalin-fixed, paraffin-embedded tissues blocks were obtain for 24 cases. In all cases, the initial diagnosis of DLBCL was confirmed. Informed consent for examination was obtained for the use of samples of all patients alive. Clinical information including age, sex, treatment and follow-up were searched from Haematology Unit records and are summarized in Table 1. Patients were divided in two main groups according to treatment: one (n=10) including patients treated with high dose chemotherapy, followed in
Table 1 Clinicopathologic features of patients Case
Age (years)
Gender
Treatment
OS (months)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
M02 M03 M05 M06 M07 M08 M10 M11 M12 M13 M15 M16 M17 M18 M19 M20
72 42 76 53 71 58 65 67 52 66 74 72 71 69 73 67
F M M M M M M F F M F M F F M M
None RT, HDC, ASCT None HDC, ASCT PT No information PT PT No information None None None PT No information PT None
1 10 6 36 3 5 34 8 3 8 0 1 24 27 3 2
17 18 19 20 21 22 23 24
M21 M22 M24 M26 M27 M28 M29 M30
61 58 42 61 69 52 51 40
M F M M M M M F
RT, HDC HDC, ASCT RT, HDC, ASCT RT, HDC, ASCT HDC RT, HDC, ASCT RT, HDC, ASCT HDC, ASCT
25 63+ 61 60+ 3 11+ 7+ 3
RT radiotherapy, HDC high dose methotrexate and cytarabine-based chemotherapy, ASCT autologous stem cell transplantation, OS overall survival, PT palliative therapy
Virchows Arch
most cases by ASCT and the other (n=14) including nontreated or palliative-treated patients.
enhancer of MGMT gene, from +57 to +150 bp (with the transcriptional initiation site as +1, refseq: NM_002412.3).
DNA extraction and bisulphite treatment
Immunohistochemical staining and scoring
Genomic DNA was extracted from three to ten representative 10-μm-thick sections using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). Bisulphite modification of 1 μg DNA was performed using the Epitect Bisulphite Kit (Qiagen) according to manufacturer instructions. Each bisulphite modification set included universal methylated and universal unmethylated DNA (Millipore, Billerica, MA, USA) as positive and negative controls, respectively, to test the effectiveness of bisulfite conversion. Modified DNAs were then used as template for PCR.
Four-micron formalin-fixed, paraffin-embedded tissue sections were dewaxed and rehydrated. Antigen retrieval was performed using DAKO’s PT-Link containing preheated (65 °C) target retrieval solution, at pH 9 (DAKO; DakoCytomation, Glostrup, Denmark), treated at 95 °C for 10 min and cooling back to 65 °C. Immunostains were performed by an automated immunostainer (Dako Autostainer; DakoCytomation), using the primary monoclonal antibody MGMT (1:10, clone MT 3.1; Thermo Scientific, Fremont, CA, USA). Standardized 3,3-diaminobenzidine (DAB) development times allowed accurate comparison of all samples. Nuclei were counterstained with Mayer's haematoxylin. Colon carcinoma tissues were used as positive control. Immunostaining was scored based on the percentage of cells that stained positively (0=50 % nuclear staining).
Real-time PCR (MethyLight) to measure CpG island methylation MethyLight analysis was performed using two sets of primers and fluorescent probes designed specifically for bisulfite-converted DNA: a methylated specific set for MGMT gene and one reference set for β-actin gene (ACTB) which was used to normalize the amount of input DNA as previously described [30, 31]. The samples were amplified by using the Epitect MethyLight Mastermix (Qiagen) according to the manufacturer's protocol in a StepOne Plus real-time PCR instrument (Applied Biosystems, Foster City, CA, USA). In each plate we generated a standard curve for MGMT and for ACTB using five serial dilutions of the positive control DNA; from the standard curves we determined the relative amounts of methylated MGMT and of ACTB amplicons for each sample. To evaluate the relative methylation level we scored a percentage of methylated reference (PMR) obtained by dividing the MGMT/ACTB ratio of a sample by the MGMT/ACTB ratio of the positive control and multiplying by 100. We considered as PMR cut-off the value of 4 that has been validated specifically for MGMT gene by previous studies [11, 13]. The definition of this value is based on the distributions of PMR values in the tested CpG island loci and on correlation of methylation with protein expression, clinical and morphological data [11, 13]. Samples showing a PMR of less than 4 were defined as unmethylated (U). Methylation-specific PCR For comparison purposes, MSP was carried out on the same set of samples following the method developed by Esteller et al. [32]. Both MSP and MethyLight assays are designed to analyze the CpGs located at the first noncoding exon and
Statistical analysis Statistical analysis was performed using Statistica 6.0 software. Since it was not possible to assume a normal distribution of data, non-parametric methods were used. Correlations between variables were evaluated using Kendall's correlation test and comparisons between groups were performed with Mann–Whitney U-test or with Kruskall–Wallis analysis of variance (ANOVA) in case of multiple comparisons. Agreement between immunohistochemical results and those of methylation analysis were analyzed with concordance correlation test and Fischer test. Overall survival rate was calculated using the Kaplan–Meier method and compared using the logrank test. Survival was measured from the time of diagnosis to the date of death or last follow-up assessment. Two-tailed p values
ORIGINAL ARTICLE
MGMT promoter methylation and correlation with protein expression in primary central nervous system lymphoma L. Toffolatti & E. Scquizzato & S. Cavallin & F. Canal & M. Scarpa & P. M. Stefani & F. Gherlinzoni & A. P. Dei Tos
Received: 30 November 2013 / Revised: 4 June 2014 / Accepted: 30 June 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract The O6-methylguanine-DNA-methyltransferase (MGMT) gene encodes for a DNA repairing enzyme of which silencing by promoter methylation is involved in brain tumorigenesis. MGMT promoter methylation represents a favorable prognostic factor and has been associated with a better response to alkylating agents in glioma and systemic lymphoma. Primary central nervous system lymphoma (PCNSL) is a rare and aggressive extranodal malignant lymphoma. The current standard of care, based on high-dose methotrexate chemotherapy, has improved prognosis but outcome remains poor for a majority of patients. Therapeutic progress in this field is conditioned by limited biological and molecular knowledge about the disease. Temozolomide has recently emerged as an alternative option for PCNSL treatment. We aimed to analyze the MGMT gene methylation status in a series of 24 PCNSLs, to investigate the relationship between methylation status of the gene and immunohistochemical expression of MGMT protein and to evaluate the possible prognostic significance of these biomarkers. Our results confirm that methylation of the MGMT gene and loss of MGMT protein are frequent events in these lymphomas (54 % of our cases) and suggest that they are gender and age related. MGMT methylation showed high correlation with loss of protein expression (concordance correlation coefficient= −0.49; Fisher exact test: p60 months in three of four patients). The prognostic significance of these molecular markers in PCNSL needs to be further studied in groups of patients treated in a homogeneous way. Keywords Primary central nervous system lymphoma . MGMT . Promoter methylation . Immunohistochemistry . Alkylating agents
Introduction The O6-methylguanine-DNA-methyltransferase (MGMT) gene encodes for a DNA repairing enzyme that removes alkyl adducts from the O6 position of guanine in DNA. MGMT is ubiquitously expressed in normal human tissues [1] while its expression in tumor cells has been shown to be considerably heterogeneous. Loss of MGMT function in human cancer is mainly due to gene silencing by hypermethylation of CpG islands located in the promoter sequence [2]. It has been demonstrated that silencing of the MGMT gene is involved in lung, brain and colorectal tumorigenesis [3, 4]. MGMT promoter methylation status and MGMT protein expression have been extensively studied and characterized in glioblastoma, anaplastic glioma and low grade glioma. In these tumors, the methylation is not only associated with a better response to alkylating agents [5–8] but also represents a favorable prognostic factor irrespective of treatment [9, 8]. In further studies, MGMT gene has been found to be methylated in some B-cell lymphomas, including diffuse large B-cell lymphoma (DLBCL) and its methylation was associated with a better response to alkylating agents [10, 11]. The most commonly used method to determine epigenetic silencing of genes is methylation-specific PCR (MSP) that can
Virchows Arch
be applied to formalin-fixed and paraffin-embedded tissue but does not allow quantitative assessment of methylation [12]. Several alternative methods have been developed, giving a quantitative value of methylation [13, 14]. Loss of MGMT protein expression detected by immunohistochemistry (IHC) has been described in a number of studies, and has shown significantly discordant results with MGMT promoter methylation as well as with survival, not only in brain tumors but also in other neoplasms [15–17]. The unreliability of IHC has been explained by tumor heterogeneity, interobserver variability, lack of cut-off values and admixture with non-neoplastic MGMT expressing cells [18]. Primary central nervous system lymphomas (PCNSLs) are extranodal malignant lymphomas arising within the brain, eyes, leptomeninges, or spinal cord in the absence of systemic lymphoma at the time of diagnosis [19]. Currently, PCNSL are estimated to account for up to 1 % of non-Hodgkin lymphomas (NHLs) and 2–3 % of all primary brain tumors [20]. Although the outcome remains poor for the majority of patients, for a selected subgroup of patients the addition to whole-brain radiotherapy (WBRT) of high-dose methotrexate (MTX) chemotherapy, followed by autologous stem cell transplantation (ASCT), has considerably improved prognosis. In such patients a 3-fold longer median overall survival compared with WBRT alone is attained and this represents the current standard of care [21–23] However, this combined treatment exposes patients, especially the elderly, to a high risk for delayed neurotoxicity and some patients cannot receive the treatment because of impaired renal function or other comorbidity [23, 24]. Since temozolomide demonstrated good penetration of the blood–brain barrier, activity in a variety of brain tumors and favorable toxicity profile, it has recently emerged as an alternative option for PCNSL, as either first-line treatment in elderly or salvage therapy after failure to high-dose MTX [25–27]. Although the majority of patients did not achieve a complete response, these studies seem to identify a subgroup of patients that show an extended progression free survival after temozolomide treatment. Recently, MGMT promoter methylation status in PCNSLs has been analyzed in order to evaluate its possible role in predicting response to temozolomide treatment [27, 28]. Another study investigated the predictive value of MGMT expression by IHC [29]. Despite the small sample size, MGMT status seems to be related with response to temozolomide in PCNSLs. Our study is aimed to analyze the methylation status of the MGMT gene in a series of 24 primary CNS lymphomas applying a standardized quantitative real-time MSP assay (MethyLight test) and comparing the results with a standard protocol of non-quantitative MSP. The relationship between the methylation status of the MGMT gene and the immunohistochemical expression of MGMT protein was investigated, and the possible prognostic significance of these biomarkers was evaluated.
Materials and methods Case selection and pathological evaluation We retrospectively searched the pathology records of the Pathology Department at the Treviso General Hospital for patients with PCNSL who underwent stereotactic biopsy or surgery from 2000 to 2011. All the original slides of available cases were reviewed by one of the authors (F.C.) for confirmation of diagnoses and evaluation of quality and quantity of neoplastic tissue. Formalin-fixed, paraffin-embedded tissues blocks were obtain for 24 cases. In all cases, the initial diagnosis of DLBCL was confirmed. Informed consent for examination was obtained for the use of samples of all patients alive. Clinical information including age, sex, treatment and follow-up were searched from Haematology Unit records and are summarized in Table 1. Patients were divided in two main groups according to treatment: one (n=10) including patients treated with high dose chemotherapy, followed in
Table 1 Clinicopathologic features of patients Case
Age (years)
Gender
Treatment
OS (months)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
M02 M03 M05 M06 M07 M08 M10 M11 M12 M13 M15 M16 M17 M18 M19 M20
72 42 76 53 71 58 65 67 52 66 74 72 71 69 73 67
F M M M M M M F F M F M F F M M
None RT, HDC, ASCT None HDC, ASCT PT No information PT PT No information None None None PT No information PT None
1 10 6 36 3 5 34 8 3 8 0 1 24 27 3 2
17 18 19 20 21 22 23 24
M21 M22 M24 M26 M27 M28 M29 M30
61 58 42 61 69 52 51 40
M F M M M M M F
RT, HDC HDC, ASCT RT, HDC, ASCT RT, HDC, ASCT HDC RT, HDC, ASCT RT, HDC, ASCT HDC, ASCT
25 63+ 61 60+ 3 11+ 7+ 3
RT radiotherapy, HDC high dose methotrexate and cytarabine-based chemotherapy, ASCT autologous stem cell transplantation, OS overall survival, PT palliative therapy
Virchows Arch
most cases by ASCT and the other (n=14) including nontreated or palliative-treated patients.
enhancer of MGMT gene, from +57 to +150 bp (with the transcriptional initiation site as +1, refseq: NM_002412.3).
DNA extraction and bisulphite treatment
Immunohistochemical staining and scoring
Genomic DNA was extracted from three to ten representative 10-μm-thick sections using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). Bisulphite modification of 1 μg DNA was performed using the Epitect Bisulphite Kit (Qiagen) according to manufacturer instructions. Each bisulphite modification set included universal methylated and universal unmethylated DNA (Millipore, Billerica, MA, USA) as positive and negative controls, respectively, to test the effectiveness of bisulfite conversion. Modified DNAs were then used as template for PCR.
Four-micron formalin-fixed, paraffin-embedded tissue sections were dewaxed and rehydrated. Antigen retrieval was performed using DAKO’s PT-Link containing preheated (65 °C) target retrieval solution, at pH 9 (DAKO; DakoCytomation, Glostrup, Denmark), treated at 95 °C for 10 min and cooling back to 65 °C. Immunostains were performed by an automated immunostainer (Dako Autostainer; DakoCytomation), using the primary monoclonal antibody MGMT (1:10, clone MT 3.1; Thermo Scientific, Fremont, CA, USA). Standardized 3,3-diaminobenzidine (DAB) development times allowed accurate comparison of all samples. Nuclei were counterstained with Mayer's haematoxylin. Colon carcinoma tissues were used as positive control. Immunostaining was scored based on the percentage of cells that stained positively (0=50 % nuclear staining).
Real-time PCR (MethyLight) to measure CpG island methylation MethyLight analysis was performed using two sets of primers and fluorescent probes designed specifically for bisulfite-converted DNA: a methylated specific set for MGMT gene and one reference set for β-actin gene (ACTB) which was used to normalize the amount of input DNA as previously described [30, 31]. The samples were amplified by using the Epitect MethyLight Mastermix (Qiagen) according to the manufacturer's protocol in a StepOne Plus real-time PCR instrument (Applied Biosystems, Foster City, CA, USA). In each plate we generated a standard curve for MGMT and for ACTB using five serial dilutions of the positive control DNA; from the standard curves we determined the relative amounts of methylated MGMT and of ACTB amplicons for each sample. To evaluate the relative methylation level we scored a percentage of methylated reference (PMR) obtained by dividing the MGMT/ACTB ratio of a sample by the MGMT/ACTB ratio of the positive control and multiplying by 100. We considered as PMR cut-off the value of 4 that has been validated specifically for MGMT gene by previous studies [11, 13]. The definition of this value is based on the distributions of PMR values in the tested CpG island loci and on correlation of methylation with protein expression, clinical and morphological data [11, 13]. Samples showing a PMR of less than 4 were defined as unmethylated (U). Methylation-specific PCR For comparison purposes, MSP was carried out on the same set of samples following the method developed by Esteller et al. [32]. Both MSP and MethyLight assays are designed to analyze the CpGs located at the first noncoding exon and
Statistical analysis Statistical analysis was performed using Statistica 6.0 software. Since it was not possible to assume a normal distribution of data, non-parametric methods were used. Correlations between variables were evaluated using Kendall's correlation test and comparisons between groups were performed with Mann–Whitney U-test or with Kruskall–Wallis analysis of variance (ANOVA) in case of multiple comparisons. Agreement between immunohistochemical results and those of methylation analysis were analyzed with concordance correlation test and Fischer test. Overall survival rate was calculated using the Kaplan–Meier method and compared using the logrank test. Survival was measured from the time of diagnosis to the date of death or last follow-up assessment. Two-tailed p values
