Hypermethylation of FHIT as a prognostic marker in nonsmall cell lung carcinoma

Authors

  • Riichiroh Maruyama M.D.,

    Corresponding author
    1. Hamon Center for Therapeutic Oncology Research, the University of Texas Southwestern Medical Center, Dallas, Texas
    2. Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan
    • Department of Thoracic Oncology, National Kyushu Cancer Center, 3-1-1, Notame, Minami-ku, Fukuoka 811-1395, Japan
    Search for more papers by this author
    • Fax: (011) 81-92-551-4585

  • Kenji Sugio M.D.,

    1. Hamon Center for Therapeutic Oncology Research, the University of Texas Southwestern Medical Center, Dallas, Texas
    2. Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan
    Search for more papers by this author
  • Ichiro Yoshino M.D.,

    1. Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan
    Search for more papers by this author
  • Yoshihiko Maehara M.D.,

    1. Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan
    Search for more papers by this author
  • Adi F. Gazdar M.D.

    1. Hamon Center for Therapeutic Oncology Research, the University of Texas Southwestern Medical Center, Dallas, Texas
    2. Department of Pathology, the University of Texas Southwestern Medical Center, Dallas, Texas
    Search for more papers by this author

Abstract

BACKGROUND

Methylation of CpG islands in the promoter and upstream coding regions has been identified as a mechanism for transcriptional inactivation of tumor suppressor genes. The purpose of the current study was to determine the correlation between the aberrant promoter methylation of multiple genes and survival in patients with nonsmall cell lung carcinoma (NSCLC).

METHODS

The methylation status of nine genes was determined in 124 surgically resected NSCLC cases using methylation-specific polymerase chain reaction.

RESULTS

The methylation frequencies of the genes tested in NSCLC specimens were 52% for E-cadherin (CDH1), 41% for RAS association domain family protein (RASSF1A), 38% for fragile histidine triad (FHIT) and adenomatous polyposis coli (APC), 27% for retinoic acid receptor beta (RARβ) and H-cadherin (CDH13), 20% for p16INK4A, 0.8% for O6-methylguanine-DNA-methyltransferase (MGMT), and 0% for glutathione S-transferase P1 (GSTP1). The survival of the patients with FHIT methylation-positive tumors was found to be significantly shorter than that for those patients with methylation-negative tumors (P = 0.03), even in those patients with International Union Against Cancer TNM Stage I or Stage II disease (P = 0.007). In contrast, there were no significant survival differences noted between the methylation-positive and methylation-negative tumors for the other genes tested. In addition, based on multivariate analyses, FHIT methylation-positive status was found to be independently associated with poor survival (P = 0.046) and disease stage (P < 0.0001).

CONCLUSIONS

The results of the current study suggest that methylation of FHIT is a useful biomarker of biologically aggressive disease in patients with NSCLC. Cancer 2004;100:1472–7. © 2004 American Cancer Society.

Lung carcinoma has been reported to be the leading cause of cancer-related deaths for men and the third leading cause for women in Japan since 1997. Lung carcinoma was reported to have killed approximately 54,000 patients in Japan in the year 2000 and was believed to be responsible for approximately 18% of all cancer-related deaths.1

DNA is methylated only at the cytosine located 5′ to guanosine in CpG dinucleotide; DNA methylation of the CpG sites in the promoter regions of genes is a frequently acquired epigenetic event in the pathogenesis of many human malignancies.2, 3 This modification has important regulatory effects that result in a loss of gene expression when it involves CpG-rich areas known as CpG islands. DNA methylation may provide an alternate pathway to gene deletion or mutation in the loss of the tumor suppressor gene function. To our knowledge, aberrant promoter methylation has to date been described for several genes in various malignancies and the spectrum of the genes involved suggests that specific tumors may have their own distinct pattern of methylation.3, 4 The methylation patterns of nonsmall cell lung carcinoma (NSCLC) have been described previously5 and hypermethylation of the RAS association domain family protein (RASSF1A),6 adenomatous polyposis coli (APC),7 and p16INK4A in Stage I adenocarcinoma of the lung8 has been reported to be associated with a poor survival. In the current study, we chose nine genes frequently silenced by aberrant methylation in lung carcinoma to determine the correlation between the aberrant promoter methylation profile and survival in patients with primary NSCLC.

MATERIALS AND METHODS

Clinical Samples

One hundred twenty-four consecutive, fresh NSCLC tissue specimens were obtained by surgical resection at the Department of Surgery and Science, Graduate School of Medical Sciences at Kyushu University between 1990–1994, after obtaining the approval of our Institutional Review Board. The patients were comprised of 79 males and 45 females who ranged in age from 25–82 years (median, 65 years) at the time of diagnosis. Twenty-seven patients had Stage IA disease, 32 patients had Stage IB disease, 5 patients had Stage IIA disease, 19 patients had Stage IIB disease, 17 patients had Stage IIIA disease, 17 patients had Stage IIIB disease, and 7 patients had Stage IV disease. All patients with Stage IV disease had ipsilateral pulmonary metastases. All patients in the current series underwent a complete resection. The histologic tumor subtypes included 76 adenocarcinomas, 40 squamous cell carcinomas, 3 large cell carcinomas, and 3 adenosquamous carcinomas. The pathologic stage of the disease was based on the TNM classification of the International Union Against Cancer (UICC).9 The histologic analysis of the tumor was based on the World Health Organization classification for cell types.10 The clinicopathologic characteristics of the patients are shown in Table 1.

Table 1. Clinicopathologic Characteristics of the Patients
ParameterNo.%
  1. UICC: International Union Against Cancer.

Median age, yrs (range)65 (25–82) 
Gender  
 Male7963.7
 Female4536.3
Histologic type  
 Adenocarcinoma7661.3
 Squamous cell carcinoma4032.3
 Large cell carcinoma32.4
 Adenosquamous carcinoma32.4
 Others21.6
Pathologic T classification  
 T13830.6
 T26250.0
 T386.5
 T41612.9
Pathologic N classification  
 N06955.6
 N12419.4
 N22923.4
 N321.6
Pathologic M classification  
 M011794.4
 M175.6
Pathologic UICC TNM stage  
 IA2721.8
 IB3225.8
 IIA54.0
 IIB1915.3
 IIIA1713.7
 IIIB1713.7
 IV75.6

The median follow-up duration of the patients was 38 months (range, 0–114 months).

Methylation-Specific Polymerase Chain Reaction Assay

Genomic DNA was isolated from frozen tissue by digestion with 100 μg/mL proteinase K followed by standard phenol-chloroform (1:1) extraction and ethanol precipitation. DNA was treated with sodium bisulfite as described previously.11 Briefly, 1 μg of genomic DNA was denatured by incubation with 0.2M NaOH for 10 minutes at 37 °C. Aliquots of 10 mM hydroquinone (30 μL) (Sigma Chemical Company, St. Louis, MO) and 3 M of sodium bisulfite (pH 5.0) (520 μL) (Sigma Chemical Company) were added and the solution was incubated at 50 °C for 16 hours. Treated DNA was purified using the Wizard DNA Purification System (Promega Corporation, Madison, WI), desulfonated with 0.3M NaOH, precipitated with ethanol, and resuspended in water. Modified DNA was stored at–70 °C until used. Methylation-specific polymerase chain reaction (MSP) was performed with primers specific for the methylated reaction as described previously.6, 11–18 Negative control samples without DNA were included for each set of PCR. PCR products were analyzed on 2% agarose gels containing ethidium bromide.

Data Analysis

Survival was calculated from the date of surgery until death or the date of last follow-up (censored). Survival was analyzed according to the Kaplan–Meier method and differences in the distribution were evaluated by means of the log-rank test.19, 20 Cox proportional hazards models were applied for multivariate analysis.21 A P value < 0.05 was defined as being statistically significant. All data were analyzed with the use of Survival Tools for StatView (Abacus Concepts, Inc., Berkeley, CA).

RESULTS

Frequency of Methylation in NSCLC

Among 124 NSCLC cases, the methylation frequencies (in descending order) were as follows: 65 (52%) for E-cadherin (CDH1), 51 (41%) for RASSF1A, 47 (38%) for fragile histidine triad (FHIT) and APC, 33 (27%) for retinoic acid receptor beta (RARβ) and H-cadherin (CDH13), 25 (20%) for p16INK4A, 1 (0.8%) for O6-methylguanine-DNA-methyltransferase (MGMT), and 0 (0%) for glutathione S-transferase P1 (GSTP1). Figure 1 illustrates representative examples of the methylation patterns in tumors of the four most frequently methylated genes: CDH1, RASSF1A, FHIT, and APC. The unmethylated form of p16INK4A, which was run as a control for DNA integrity, was present in all samples.

Figure 1.

Representative examples of methylation-specific polymerase chain reaction analyses of the methylated form (M) of four genes frequently methylated in nonsmall cell lung carcinoma: E-cadherin (CDH1), RAS association domain family protein (RASSF1A), fragile histidine triad (FHIT), and adenomatous polyposis coli (APC). The amplification of the unmethylated form of p16INK4A (p16U) was used as a control for DNA integrity. T: tumor samples 1–10; N: negative control; bp: base pairs.

Correlation between Methylation and Prognosis

The 5-year survival rates with regard to the methylation status of the genes tested were as follows: FHIT, 31.9% for methylation-positive patients and 51.4% for methylation-negative patients; RASSF1A, 37.7% for methylation-positive patients and 45.2% for methylation-negative patients; CDH13, 41.9% for methylation-positive patients and 42.4% for methylation-negative patients; APC, 38.8% for methylation-positive patients and 44.2% for methylation-negative patients; RARβ, 48.3% for methylation-positive patients and 39.9% for methylation-negative patients; CDH1, 42.4% for methylation-positive patients and 42.0% for methylation-negative patients; and p16INK4A, 47.7% for methylation-positive patients and 41.0% for methylation-negative patients. The patients who were otherwise diagnosed with Stage I or II disease were found to have a significantly shorter survival when their methylation FHIT status was positive compared with those patients with Stage I or II disease who had a negative FHIT methylation status, as shown in Figure 2A (P = 0.007). Patients with all stages of disease who had methylation-positive FHIT status were found to have significantly shorter survivals compared with the patients with methylation-negative FHIT status, as shown in Figure 2B (P = 0.03). In contrast, there were no significant survival differences noted between the methylation-positive and methylation-negative tumors for the other genes tested.

Figure 2.

Correlation of the methylation status of FHIT and patient survival using the Kaplan–Meier method. (A) The survival curves in the patients with International Union Against Cancer TNM Stage I or Stage II disease. (B) The survival curves of the patients with all stages of disease. U: unmethylated case; M: methylated case.

In a multivariate analysis model that included age, gender, histologic type (large cell carcinoma and adenosquamous cell carcinoma were included in the “others” category), the stage of disease, and the methylation status of frequently methylated (≥ 20%) genes (CDH1, RASSF1A, FHIT, APC, RARβ, CDH13, and p16INK4A), positive FHIT methylation status was found to be an independent prognostic factor (P = 0.046), as was the stage of disease (P < 0.0001) (Table 2).

Table 2. Multivariate Statistics of Survival
VariableHazards ratio95% CIP value
  1. 95% CI: 95% confidence interval; adeno: adenocarcinoma; UICC: International Union Against Cancer.

Age (yrs)   
 ≥ 65 vs. < 651.400.84–2.330.20
Gender   
 Female vs. male0.760.45–1.310.32
Histologic type  0.36
 Squamous vs. adeno1.260.70–2.260.44
 Others vs. adeno0.580.19–1.810.35
Pathologic UICC TNM stage  < 0.0001
 Stage II vs. Stage I3.431.82–6.460.0001
 Stage III vs. Stage I4.212.32–7.64< 0.0001
 Stage IV vs. Stage I4.551.46–14.20.009
FHIT methylation1.761.01–3.080.046
RASSF1A methylation1.270.77–2.100.35
CDH 13 methylation1.170.65–2.090.61
APC methylation0.960.57–1.620.87
p16 methylation0.810.43–1.520.52
CDH 1 methylation0.730.43–1.260.26
RARβ methylation0.580.32–1.060.078

DISCUSSION

Previous studies have described the importance of DNA methylation in human tumors while focusing on regions of the genome that might have functional significance resulting from the abolition of all gene activity. Although the majority of individual tumors have several, perhaps hundreds, of methylated genes,3 the methylation profiles of individual tumor types are characteristic.3–5 We determined the methylation profile of NSCLC cases by testing a panel of nine genes that were studied extensively in many tumor types by the widely utilized MSP assay. Frequent methylation (≥ 20%) was noted for 7 of the 9 genes tested: RASSF1A (41%), APC and FHIT (38%), RARβ (27%), p16INK4A (20%), and 2 members of the cadherin family (CDH1 [52%] and CDH13 [27%]).

The FHIT gene, located in the chromosome region 3p14.2, undergoes frequent allele loss (loss of heterozygosity) and occasional homozygous deletions in lung carcinoma, whereas to our knowledge, point mutations are rare.22–24 The loss of FHIT protein expression also is reported to occur frequently in lung carcinoma25–28 and, based on an immunohistochemical analysis, a reduction in FHIT protein expression was reported to be associated with a poorer survival in patients with Stage I NSCLC.26 Recently published studies concluded that methylation of FHIT is a frequent event in NSCLC12 and it also is associated with a shortened survival in patients with bladder carcinoma.29 In the current study, the promoter methylation of FHIT was detected in 38% of NSCLC cases and was associated with decreased survival, even in those patients with Stage I or Stage II disease.

Several members of the cadherin gene family, including CDH1 and CDH13, are located on chromosome 16q, a region of frequent allelic loss in multiple tumor types.30CDH1 is inactivated by methylation in several tumor types, including carcinomas of the lung, breast, and esophagus.4, 5 We found the methylation of CDH1 to be present in 52% of NSCLC cases in the current study. The findings of the current study indicate the methylation rate in CDH1 to be lower than in the latter study.5 Inactivation of CDH13 by promoter methylation has been described in both lung and breast carcinomas.18, 31 We found the methylation of CDH13 to be present in 27% of NSCLC cases in the current study.

Our gene panel included RASSF1, a recently identified putative tumor suppressor gene. There are two major RASSF1 gene products: RASSF1A and RASSF1C. The selective promoter methylation of the RASSF1A promoter, but not of RASSF1C, is reported to be frequent in both small cell lung carcinoma and NSCLC and in breast carcinoma6, 32 and the methylation of RASSF1A also has been reported to be associated with poor survival.6 However, no significant association between RASSF1A methylation and survival was observed in either the current study or in a previous large, multinational study.33

Inactivation of the APC gene is a frequent occurrence in colorectal and other gastrointestinal carcinomas, usually by truncating mutations.34 An alternative method of inactivation of the gene in some gastrointestinal tumors is by promoter methylation.35 Virmani et al. reported that selective methylation and silencing of the 1A promoter and its specific products is frequently found in lung carcinoma and breast carcinoma.36 A recently published study reported a significant association between APC promoter methylation and a shorter survival in patients with NSCLC.7 No significant association between methylation of APC and survival was observed in the current study.

The p16INK4A protein inhibits cyclin-dependent kinase 4, a key regulator of disease progression through the G1 phase of the cell cycle. A recently published study reported the methylation of p16INK4A to be associated with exposure to tobacco smoke in NSCLC as well as a significant association between p16INK4A methylation silencing in Stage I adenocarcinomas.8 According to the current study data, no significant association between p16INK4A methylation and patient outcome was observed, even in those patients with early-stage adenocarcinoma.

None of the tumors in the current study demonstrated the promoter methylation of MGMT and GSTP1, with the exception of one case of MGMT. The findings of the current study regarding the methylation rates in MGMT were lower than those of most other reports,4, 5, 15 although the primers used in the latter study covered the same sites in the MGMT promoter as did ours. As a result, there is evidence of geographic differences in the methylation of these genes.33

Although the findings of the current study still need to be confirmed in a larger series, they suggest that the methylation of FHIT is a useful biomarker of biologically aggressive disease in patients with NSCLC.

Acknowledgements

The authors thank Dr. Brian T. Quinn for critical comments on the article.

Ancillary