Abstract
Background: An association between restriction fragment length polymorphism (RFLP) of known oncogenes and a predisposition to develop cancer has been postulated. Our aim was to test the hypothesis that there was an association between the L-MYC S allele in oral squamous cell carcinoma (OSCC) and a predisposition for the disease. Patients and Methods: The distribution of L-MYC polymorphism in 80 patients with OSCC was determined by polymerase chain reaction-based RFLP and compared with that of 60 healthy controls. Results: There was no significant difference between patients with OSCC and healthy controls. Patients with the L-MYC S allele and a positive family history of cancer were found to be 1.74 times more at risk for OSCC than those with any other genotype (95% confidence interval=0.88-3.45). Moreover, tumor recurrence was higher among individuals carrying a L-MYC S allele than those with any other allele type. Conclusion: L-MYC polymorphism was not a significant marker for predicting susceptibility to OSCC in this population but may be a useful marker for identifying patient susceptibility to tumor recurrence and to developing OSCC, especially in individuals having a family history of cancer.
Oral squamous cell carcinoma (OSCC) is the most common cancer of the maxillofacial region (1, 2). According to the worldwide estimations, of all cancers, oral cancer accounts for 5.6% in men and 3.2% in women (3). Although this rate is low compared with major cancer types, the 5-year survival rates are not promising at present. Etiology of oral cancer is multi-factorial because a genetic relation (4), as well as other defined habits such as tobacco and alcohol consumption (5-8), is involved. In an effort to find a suitable genetic marker for the assessment of individual risk of cancer, polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) analysis of known oncogenes, such as the L-MYC gene, has been carried out in a variety of cancer patients (9, 10). L-MYC is a member of the MYC gene family which was initially identified in a human small cell lung cancer cell line due to its structural similarity to c- and N-myc (11, 12). L-MYC exhibits two-allele polymorphism, with 10 kb (L) and 6.6 kb (S) alleles revealed after EcoRI restriction enzyme digestion (11). Previous studies have reported that patients having the S allele showed a much higher incidence of metastasis in renal (13), oral (14), gastric and breast (15), colorectal (16), esophageal (17) and lung cancer (18). On the other hand, investigators have failed to find its association between Caucasian lung cancer (19-21), colorectal cancer (22), non-Hodgkin lymphoma (23) and acute lymphocytic leukemia (24).
In the present study, the L-MYC gene was analyzed in a group of Turkish OSCC patients using RFLP in order to investigate whether there is any relation between the L-MYC S allele and tumor node metastasis (TNM) classification.
Patients and Methods
Study participants. In this study, samples were obtained from İstanbul University Medical Faculty Department of Otorhinolaryngology and Department of Radiation Oncology. L-MYC gene polymorphism was studied in 80 OSCC patients (mean age 54.9±13.21 years) and 60 age-matched healthy individuals without any disease history as a control group (mean age 50.82±11.11 years). Detailed medical history, oral examination and pathological diagnosis were performed for all patients in the study. All OSCC patients were classified by TNM staging system. A standardized questionnaire was administered to collect data concerning age, sex, smoking history, alcohol consumption, tumor stage and histological grade of tumor for OSCC patients.
Distribution of L-MYC genotypes and allele frequencies in OSCC patient and control groups.
DNA isolation. Blood specimens from all individuals were collected in tubes containing EDTA. DNA was isolated from blood leukocytes in 10 ml EDTA by the method developed by Miller et al. based on sodium dodecyl sulphate lysis, ammonium acetate extraction and ethanol precipitation (25).
Polymerase chain reaction (PCR) for L-MYC oncogene. Template DNA (0.5-1.0 μg) was used in a PCR under sterile conditions with 100 ng of primer for the reaction. The forward primer was 5′-AGT TCA CTC ACA GGC CAC AT-3′ and the reverse primer was 5′ TGC ATA TCA GGA AGC TTG AG-3′ in a volume of 50 μl containing 3 mM MgCl2, 50 mM KCl, 10 mM Tris-HCl (pH:8.4), 0.5 mM of each dNTP (MBI Fermentas, Vilnius, Lithuania) and 1 unit of Taq polymerase (MBI Fermentas). Amplification was carried out in a DNA Thermal Cycler (MJ Research Techn) for 30 cycles with denaturation steps at 94°C for 30 seconds, annealing at 50°C for a minute, and extension at 74°C for 1 minute. The amplified 267 bp PCR product was directly digested in the process catalyzed by restriction enzyme EcoRI (MBI Fermentas). The digested DNAs were separated on 2% agarose gel in 1×Trisborate EDTA buffer and were visualized by ethidium bromide staining. L-MYC oncogene polymorphism was typed for visualization under ultraviolet light and was photographed with a Polaroid camera (19).
Statistical analysis. Statistical analysis was performed using GraphPad Prism V.3 program (BioPortfolio Inc., CA, USA) including Chi-square (χ2) test and odds ratio for the qualitative data.
Results
The distribution of L-MYC polymorphism in 80 patients with OSCC was determined by PCR-RFLP and compared with that of 60 healthy controls. The three genotypes were identified as: LL homozygote appearing in a 267-base pair (bp) fragment, LS heterozygote with 267, 142 and 125-bp fragments, and SS homozygote with 142 and 125-bp fragments.
The L-MYC oncogene genotypes and allele frequencies for OSCC and healthy control groups are shown in Table I. The relative ratios of the LL, LS and SS genotypes were not significantly different between OSCC patients and healthy controls (χ2=0.98, p=0.611). The allele frequencies of L and S in oral cancer patients were 0.575 and 0.425 while those in healthy controls were 0.592 and 0.408, respectively. There were no significant differences between OSCC patients and healthy controls in term of L-MYC allelic frequencies (χ2=0.078, p=0.779). Data on age, sex and smoking status of OSCC patients were also correlated with the L-MYC genotypes and have been summarized in Table II. The age of OSCC patients was divided between <45 and >45 years; a parameter set in other papers on etiology of oral cancer (26, 27). No statistical difference between age group and L-MYC genotype was found. Data for the healthy control group on the distribution of genotypes according to sex showed a slightly higher number of women with the SS genotype as compared with men. The frequencies of LL, LS and SS genotypes in women in the healthy control group were 0.278 (n=10), 0.472 (n=17), and 0.250 (n=9), while those in men were 0.500 (n=12), 0.417 (n=10) and 0.083 (n=2), respectively. There were no significant differences between the different groups (p=0.121), but the relative likelihood of women having the SS genotype was 3 times higher than that of men (OR=3, 95% confidence interval (CI)=0.7-12.7). Like the control group data, the number of women OSCC patients with SS genotypes was higher than men, also having no statistical difference. To assess the smoking status of OSCC patients, they were asked in the questionnaire of the number of cigarettes smoked per day and the duration of their smoking habit. Fifty-three out of 80 OSCC patients were smokers. Past smokers were those who had given up smoking after the diagnosis of their disease. Twenty-six out of 53 smokers (62.3%) smoked more than 20 cigarettes per day, and 37 patients had been smoking for more than 20 years. The L-MYC genotype distribution of the smokers did not differ statistically from that of non-smokers in respect of time and the amount of smoking. Abstinence from alcohol use as well as smoking was considered as a major factor in avoiding OSCC (6). According to the British Health Council, alcohol consumption affects human health if more than 21 units a week are consumed by men and 14 units by women (1 unit=1 small glass of wine) (5). In our study, we also considered this information to decide whether the OSCC patients were social drinkers or not. Social drinkers and individuals that do not consume alcohol were considered as alcohol (-) and only individuals who consumed more alcohol units than mentioned above were considered as alcohol (+). Twenty-five (31.3%) of the OSCC patients were found to be alcohol users but the L-MYC gene distribution of these patients did not show any significant difference due to their alcohol behavior.
Distribution of L-MYC genotypes according to the data on OSCC patients.
Clinical pattern of OSCC patients in correlation with L-MYC genotypes.
The relationship between L-MYC genotype and the level of differentiation, nodal metastasis, size of primary tumor and TNM staging (28) and also recurrence of OSCC in 32 patients who had follow-up data for more than 24 months are summarized in Table III. Twenty-six patients had well-differentiated tumors, whereas 36 had moderate and 17 had poorly differentiated tumors. We excluded one patient's data because the patient had undergone two surgeries and demonstrated two different differentiation levels (one moderate and one poor) as revealed by the pathology. The presence of SS or LS genotype (the genotypes with an S allele) was not significantly related to the level of differentiation. The primary tumor size of OSCC patients was also tested for correlation with the L-MYC genotype. Fifty-two (65%) patients had tumors <4 cm. Although the p-value was >0.05, a slight increase of the presence of L-MYC S allele in small sized tumors (OR=2.2, 95% CI=0.67-7.22) was found. Thirty-five (43.8%) OSCC patients had nodal metastasis. The L-MYC genotype was not significantly related to tumor size in these groups. According to TNM classification, 33 (41.3%) of the patients were either grade I or grade II and the rest 47 (58.7%) were grade III or IV. No correlation was observed between nodal metastasis and grades of OSCC and the presence of the L-MYC S allele. Follow-up data of at least 24-months were available for 32 patients: twenty-six (81.3%) patients did not show recurrence within this period, while 6 (18.7%) of them relapsed. The analysis of L-MYC genotypes demonstrated that S allele carriers had 3.12 times more recurrence risk than L allele carriers (95% CI=0.32-30.79).
The distribution of L-MYC genotypes and allelic frequencies according to family history in OSCC patients.
The family history regarding every kind of cancer was analyzed in all the OSCC patients. Patients with the L-MYC S allele and positive family history were found to be 1.74 times more at risk for OSCC than those with any other genotype (95% CI=0.88-3.45) (Table IV).
Discussion
The study of Kawashima et al. (18) was the first report on L-MYC RFLP of Japanese lung cancer patients. Since then, many studies have demonstrated an association between the S allele of the L-MYC gene and susceptibility to certain types of cancers. This is also true for the clinical pattern of various cancer types which was clearly summarized in the meta-analysis by Spinola et al. (29). Still the results were conflicting even in the studies on the same cancer type.
In the case of gastric cancer, two studies (one Japanese, one Turkish) found a close relation between L-MYC RFLP and susceptibility to gastric cancer (30, 31), whereas the other two studies (one Japanese, one Russian) failed to find any association (15, 32). These differences were mainly attributed to the different populations on which the studies were conducted and also the method used.
In a study by Weston et al. (21), the distribution of L and S alleles among different ethnic groups was investigated. The ratios of L alleles to S alleles in Caucasians and African-American populations were 0.509:0.491 and 0.277:0.763, respectively. In that report, a significant difference between distributions of alleles was observed. When the frequency of the S allele in the controls of different populations was compared, the only statistical difference observed was again in the African-American group compared to Caucasians. The Japanese and Thai frequencies were similar to Caucasians, whereas the Indian and Russian frequencies were slightly elevated.
In our study, the ratios of the L allele to the S allele in controls and OSCC patients were 0.617:0.383 and 0.575:0.425, respectively. We did not find a significant difference in allele frequencies of the L-MYC gene between controls and OSCC patients (p=0.779).
The OSCC patients with the SS genotype showed a slight risk of predisposition for the disease compared to those with the LL genotype (OR=2.2; 95% CI=0.48-3.08). In a previous Indian study on L-MYC RFLP in oral cancer, the distribution of genotypes in the patient group was 30.4% SS, 48.6% LS and 21% LL and in controls, it was 21.7% SS, 48.5% LS and 29.7% LL, respectively (14). Although the distribution of percentages in our study is not similar to that of the Indian oral cancer study, in both studies there was no statistically significant difference found between genotype distribution and predisposition for oral cancer.
Most of the previous studies in the literature used no age- or sex-matched controls (14, 23, 33-35). Recently published studies have placed greater emphasis on the importance of age- and sex-matched controls (17, 31, 36-38). Shih et al. (38) have attributed the conflicting data in the literature to the results of studies without age and sex-matched controls as well as differences in methodology and ethnicity.
Female oral cancer patients tended to have higher frequency of the SS genotype (OR=2.3, 95% CI=1.06-4.98) compared to male patients and similar findings were also noticed in the control group (OR=3, 95% CI=0.7-12.7). Togo et al. (35) found a slight increase of S allele in male controls. Shih et al. (38) reported that control group genotypes showed no difference according to sex, where as the SS genotype was slightly more frequent in females with lung cancer (OR=1.18, 95% CI=0.49-2.84). These results suggest that there are no marked differences in genotype distribution and predisposition for oral cancer based on sex, although sex predilection of the SS genotype in oral cancer tends to draw some attention.
Age also did not prove to be a marker of differences in L-MYC genotypes. The conspicuous finding was an increase of homozygote genotypes (both SS and LL) with aging in both the oral cancer and control groups. These data correspond with the findings of Togo et al. (35) in comparing middle-aged healthy individuals with tumor-free elderly ones.
In the present study, we found no correlation of genotypes related to age, smoking history, alcohol use, TNM stage or nodal metastasis. Similarly, the study of Saranath et al. (14) also failed to find any association between L-MYC allelotype, TNM stage and nodal metastasis in Indian oral cancer patients.
Many investigators have analyzed L-MYC RFLP types in different cancer studies. Although some studies have reported that SS or LS genotypes were correlated with lymph node metastasis in stomach (39), renal (13) and rarely lung cancer (18, 33, 40), no correlation was found with renal cell carcinoma (41), colorectal (22), prostate (42), bladder (43) or some cases of lung cancer (19, 20, 44).
In the present study, we found an inverse correlation between the presence of L-MYC S allele and tumor size (OR=2.2, 95% CI=0.67-7.22). In the study by Saranath et al. (14), positive correlation was found with the tumor size (p<0.025). However, this contradiction is not surprising when the number of cases with primary tumors less than 4 cm diameter is taken into account in both the studies. In our study, there were 52 cases (52/80) with primary tumor less than 4 cm, while in the Indian study, there were only 22 cases (22/76) with similarly sized tumors. We believe that if the proportion of the subgroups was alike, the results might be clearer in terms of racial differences and other factors such as age and sex which have not been clearly mentioned in the previous work.
The grades of differentiation were categorized as either poor, moderate or well in the histopathological examination of the tumor tissues. We found no significant differences in L-MYC genotype distribution between the three groups (p=0.954). In the Indian study, the poor and moderate differentiations were grouped as one and found to be correlated with the S allele (p<0.001), as compared with the well-differentiated cancer group of patients (14). In our patient group, the number of well-differentiated tumors was 26 out of 80, similar to that of the Indian study group with 25 out of 76. We had 36 moderate and 17 poorly differentiated tumors which we could not compare with the Indian study since this was not clearly mentioned and only a total of 51 was given for poor + moderately differentiated tumors. However, our findings are in parallel with renal (13) and lung cancer (18) studies which found no association with the grade of tumor cell differentiation.
We evaluated tumor recurrence in 32 patients who had a follow-up period of more than 24 months. There was no significant difference (p=0.311) between allelic groups, but the S allele frequency was elevated in the recurrent subgroup (OR=3.12, 95% CI=0.32-30.79). In the previous literature, we found conflicting data on L-MYC genotypes and recurrent disease association. Even though the report of Kondratieva et al. (45) reveals a weak association in relapse of gliomas and L-MYC genotypes, Saranath et al. (14), as well as Presti et al. (41), failed to find any association between recurrent disease and L-MYC genotype in oral cancer and renal cell carcinoma patients. Our findings suggest that the recurrence of oral cancer might be associated with the L-MYC S allele but still further studies in a larger series of patients are needed to verify the relationship.
We found a weak tendency toward a higher frequency of S-allele in a cohort of patients with a positive family history of cancer compared to patients with no known first-degree relatives with cancer (p=0.152, OR=1.74, 95% CI=0.88-3.45). These data were confirmed by the Norwegian lung cancer study by Tefre et al. (20) while another weak association has been reported by İsbir et al. (46) in the case of breast cancer patients (OR=1.05, 95% CI=1.053-7.453). Some other studies on stomach and lung cancer (31, 44), where a family history of cancer was also investigated, failed to find an association with the S allele. In epidemiological studies on oral cancer, a positive family history of cancer in first degree relatives was found to be associated with 4.3-75% of females and 1-59% of males (26, 27). These data are important for oral cancer predisposition and suggests that there is an unknown gene or combination of genes that makes an individual prone to oral cancer. Here in our study, we suggest that S allele carriers with a positive family history of cancer might be more prone to oral cancer compared to other cancer types.
Our results suggest that L-MYC polymorphism is not a significant marker for assessing susceptibility to OSCC in this Turkish population but may be a useful marker for identifying patients' susceptibility to tumor recurrence and developing OSCC in individuals having a family history of cancer.
Acknowledgements
The present work was supported by the Research Fund of Istanbul University (Project No T-146/06032003) and presented as a poster presentation with the support of the Research Fund of Istanbul University (Project No: UDP-804/05072006) at the 8th Biennial Meeting of the European Association of Oral Medicine where Dr. Kivanç Bektaş-Kayhan received a ‘Young Scientist Award’ for her work.
- Received January 3, 2009.
- Revision received April 27, 2009.
- Accepted May 6, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.