Quantitative O⁶-Methylguanine DNA Methyltransferase Methylation Analysis in Curatively Resected Non-Small Cell Lung Cancer: Associations with Clinical Outcome

Lung cancer is one of the most common malignancies in the world and is the leading cause of cancer-related deaths both in the United States and worldwide (1, 2). Radical surgery offers the only chance for cure in patients with NSCLC,² but despite improvements in the detection and treatment of lung cancer in the past 2 decades, the 5-year survival rate remains <15% (1). Current clinical means cannot predict whether a patient may be cured by surgical treatment alone or will require additional and more aggressive treatment to improve the long-term survival. Therefore, it is desirable that novel and clinically applicable strategies be developed to augment the current NSCLC staging system for better classification of curatively resectable disease.

Hypermethylation of CpG islands within the promoter and 5′ regions of genes has become established as an important epigenetic mechanism for suppression of gene expression (3, 4, 5). Hypermethylation of tumor suppressor genes such as the APC gene (6, 7, 8), the p16/INK4a gene (9, 10), the DNA mismatch repair gene hMLH1 (11), the E-cadherin gene (12), the DAPK gene (13), and the DNA repair gene MGMT (14, 15) is a fundamental process involved in the development of many cancers, including NSCLC (13, 14, 15). Recent studies report an association between clinical outcome and hypermethylation of distinctive genes, for example DAPK (13) and APC (16, 17), in patients with NSCLC, and APC in patients with adenocarcinoma of the esophagus (7). Furthermore, methylation of the MGMT gene has been identified as a predictive marker for response to carmustine therapy in gliomas (18) and as a prognostic marker in patients with diffuse large B-cell lymphoma (19). These findings indicate that aberrant promoter methylation might be suitable for the approach for using DNA-based markers as prognostic tools in lung cancer and other human malignancies.

To determine the frequency and clinical relevance of MGMT promoter hypermethylation in NSCLC, we performed a methylation-specific real-time quantitative PCR (Taqman; Ref. 20) on surgically removed tumor specimens and adjacent benign tissue from 90 patients with NSCLC.

The study specimens were paired tumor and normal lung tissues from 90 patients with NSCLC that were available from a previous prospective clinical trial of 103 consecutive patients with completely resected (R0-resection) NSCLC. There were 68 (76%) men and 22 (24%) women, with a median age of 63.3 years (range, 34–82 years). Forty-three (48%) patients had squamous cell carcinomas, 32 (35%) had adenocarcinomas, and 15 (17%) had large cell carcinomas. The primary tumors were graded histopathologically as well-differentiated (G1, one patient), moderately differentiated (G2, 18 patients), and poorly differentiated (G3, 71 patients). Tumor staging was performed according to the International Union Against Cancer Tumor-Node-Metastasis classification (21): 44 (49%) had stage I tumors, 19 (21%) had stage II tumors, and 27 (30%) had stage IIIa tumors. All of the tumors were radically removed by lobectomy (n = 57), bilobectomy (n = 12), pneumonectomy (n = 11), and extended pneumonectomy (n = 10) including mediastinal lymphadenectomy for all of the procedures. Patients with histopathological stage IIIa tumors received postoperative radiotherapy. The median follow-up was 7.2 years (range, 0.3–8.8 years), and no patient was lost to follow-up. Informed consent was obtained from all of the patients. Tissue for DNA promoter methylation analysis was obtained immediately after lung resection before starting mediastinal lymphadenectomy, and was immediately frozen in liquid nitrogen and stored at −80°C. Tissue was analyzed from the following two locations: tumor and uninvolved lung tissue taken from the greatest distance to the tumor. Six-μm frozen sections were taken from blocks of tumor tissue, and starting with the first section every fifth was routinely stained with H&E and histopathologically evaluated. Sections were pooled for analysis from histological normal lung tissues and areas of estimated 75% malignant cells for matching tumor analysis. Ten histologically normal lung tissue specimens, obtained at surgery from patients with no evidence of cancer were used as a control group. Nine (90%) men and 1 (10%) woman, with a median age of 46 years (range, 1–64) were included in this group.

Genomic DNA was isolated from frozen tissue by standard methods of proteinase K digestion and phenol-chloroform extraction using the GenomicPrep Cells and Tissue Isolation kit (Amersham Pharmacia Biotech Inc., Piscataway, NJ) according to the instructions of the manufacturer.

Sodium bisulfite conversion of genomic DNA was performed as described previously (22). The protocol was modified according to Eads et al. (23).

After sodium bisulfite conversion, quantitative methylation analysis was performed using a fluorescence based real-time detection method (TaqMan), as described previously (20, 23, 24). The primers and probes are listed in Table 1. MGMT primers and probes were specifically designed to amplify methylated, bisulfite-converted DNA, and contain 12 CpG islands. MYOD1 was used as a reference gene to normalize for input DNA. The reference primers and the probe were designed in a region of the MYOD1 gene that lacks any CpG dinucleotides to allow for equal amplification, regardless of any methylation levels. Parallel TaqMan PCR reactions were performed for methylated MGMT and the MYOD1 reference gene for each sample. Each reaction was repeated at least once. The ratio between the values obtained was used as a measure for the degree of methylation. A MGMT methylation-positive human lung cell line L132 was used as a positive control and for constructing the calibration curve on each plate.

Associations between two related variables were tested by using the Wilcoxon signed rank test. Associations between categorical variables were tested with the Fisher’s exact test. Survival was estimated according to Kaplan and Meier (25), with the stratified log-rank test (in which patients were stratified according to stage) used to measure the strength of the grouping. Multivariate analysis was performed with the Cox proportional hazard regression model (26). All of the reported Ps are based on two-sided tests.

MGMT promoter hypermethylation was detectable by quantitative real-time PCR in 16 of 90 (17.8%) normal lung, 34 of 90 (37.8%) matching tumor samples of NSCLC patients, and none of 10 (0%) normal lung specimens of the control group without lung cancer. Methylation in tumors but not surrounding nontumor tissue was observed in 18 of 90 (20%) cases, but there was no case in which methylation was seen in normal tissue but not in corresponding tumor tissue. The mean MGMT promoter methylation, expressed as the ratio between methylated MGMT DNA and MYOD1 PCR product, was 0.03 (range, 0.00–0.55) in normal lung and 2.08 (range, 0.00–134.9) in tumor tissue (P <0.001).

Table 2 shows associations between various clinicopathological data and MGMT promoter methylation in tumor tissues of patients with undetectable (negative) MGMT methylation and detectable (positive) MGMT methylation. No statistically significant differences were found.

With a median follow-up of 7.2 years for the 90 patients analyzed in this study, the median survival was 4.3 (95% CI, 2.8–8.8) years. Median and 5-year survival rates depending on various clinical variables and MGMT methylation status are summarized in Table 3. To determine whether there was any prognostic significance attached to differences in the MGMT promoter methylation status, we compared the clinical outcomes for patients that were negative for MGMT promoter methylation (n = 56) and those that were positive for MGMT promoter methylation (n = 34) in tumor tissue (Fig. 1). The median survival was not reached in the group that was lacking MGMT promoter methylation, whereas it was 2.6 (95% CI, 1.6–3.6) years in the group that was positive for MGMT promoter methylation (P = 0.017; stratified log-rank test). The importance of MGMT methylation as a prognostic factor was next determined by the Cox proportional hazards model analysis. The logistic regression model included parameters gender, age, histopathological tumor type, tumor stage, grade of differentiation of the primary tumor, and MGMT methylation status. Only tumor stage (P < 0.001) and MGMT methylation status (P = 0.030) were of independent prognostic importance (Table 4).

MGMT is a DNA repair protein that removes mutagenic and cytotoxic adducts from the O⁶-guanine in DNA (27). Alkylation of DNA at the O⁶ position of guanine is an important step in the formation of mutations in cancer, primarily because of the tendency of O⁶-methylguanine to pair with thymine during replication, resulting in a conversion of guanine-cytosine to adenine-thymine pairs in DNA (28, 29). Loss of MGMT expression is not commonly because of deletion or rearrangement of the gene (30, 31, 32), but rather methylation of CpG islands in the MGMT promoter region is associated with silencing of the MGMT gene in cell lines (33, 34, 35), and is associated with loss of protein expression in primary human neoplasia (15). Furthermore, in vitro treatment with demethylating agents restores the expression of the MGMT gene of such cell lines (15, 34). It was reported recently for the first time that the methylation status of the MGMT promoter is associated with clinical outcome in human cancer. Esteller et al. (18) showed that inactivation of the MGMT gene by promoter methylation is a predictor of overall survival and response to alkylating agents in patients with gliomas, additionally supporting the promise of methylated markers as prognostic tools in human cancers.

Methylation of the MGMT gene in NSCLC has been reported previously. Esteller et al. (15) found that MGMT was hypermethylated in 10 of 34 (29%) primary NSCLC tumors, and Zöchbauer-Müller et al. (36) detected MGMT methylation in 21% of NSCLC, indicating that hypermethylation of the MGMT promoter is a common event in this disease. In contrast to the present study in which a strong association between MGMT promoter methylation and survival in patients with NSCLC was present, MGMT promoter methylation was not associated with clinical outcome in either of these investigations. There are several possible explanations for these discordant findings. Firstly, the primers used in the latter study did not cover the same sites in the MGMT promoter as did ours. Secondly, MGMT hypermethylation was determined using the conventional qualitative methylation-specific PCR, whereas a quantitative real-time PCR (Taqman) technique was used in our study. The reported 10-fold greater sensitivity of the Taqman method may account for the obtained higher frequency of methylated MGMT (24). Thirdly, the populations analyzed in these studies are considerably different. Esteller et al. (15) investigated a total of 34 patients with NSCLC, and Zöchbauer-Müller et al. (36) analyzed NSCLC stage I to IIIb, whereas only patients with curatively (R0 resection) NSCLC stage I to IIIa were included in our study. Our findings complement the results of other studies showing an association between survival of NSCLC patients and other epigenetic alterations, such as methylation of DAPK (13) in stage I NSCLC, methylated E-cadherin gene in NSCLC stage I to IIIb (36), and methylated APC in curatively resected NSCLC (16, 17).

We also observed MGMT methylation in an appreciable percentage of nonmalignant tissues taken from cancer patients at the time of surgery, although at a significantly lower level than in the tumor tissue. Although this finding was somewhat unexpected, hypermethylation of RARβ, TIMP-3, DAPK, and p14 in nonmalignant lung tissue of patients with NSCLC has been reported before (36). It is possible that this noncancerous tissue, although it appeared to be histologically normal, is abnormal because of environmental factors such as exposure to cigarette smoke. Alternatively, molecular events of tumorigenesis (such as methylation of some tumor suppressor genes) may have already occurred in these tissues. Indeed, the observation that methylated MGMT was only present in noncancerous tissues when there was detectable methylation of MGMT in tumor tissue and the lack of any MGMT methylation in normal lung tissue from patients without lung cancer suggest that this methylation event is part of the process of malignant transformation, and, thus, a tumor-specific phenomenon in lung cancer.

Our study demonstrates that methylation of the MGMT promoter might be an important biomarker for biologically more aggressive disease in patients with NSCLC. These findings add another step toward the development of a model for molecular classification of NSCLC and suggest that quantitation of MGMT promoter hypermethylation might have value in identifying NSCLC patients at high risk of early disease recurrence after surgery, and in selecting patients who will benefit from intensive adjuvant therapy. On the basis of the results presented, additional studies are warranted to investigate the prognostic value of MGMT methylation in NSCLC and other human malignancies.