An increasing number of studies indicate that reduced DNA-repair capacity is associated with increased cancer risk. Using a functional assay for the removal of the oxidative DNA lesion 8-oxoguanine by the DNA-repair enzyme 8-oxoguanine DNA glycosylase 1 (OGG1), we have previously shown that reduced OGG activity is a risk factor in lung cancer. Here, we report that OGG activity in peripheral blood mononuclear cells from 37 cases with squamous cell carcinoma of the head and neck (SCCHN) was significantly lower than in 93 control subjects, frequency matched for age and gender. Retesting of OGG activity 3 to 4 years after diagnosis and successful treatment of 18 individuals who recovered from the disease showed that OGG activity values were similar to those determined at diagnosis, suggesting that reduced OGG activity in case patients was not caused by the disease. Logistic regression analysis indicated that the adjusted odds ratio (OR) associated with a unit decrease in OGG activity was statistically significantly increased [OR, 2.3; 95% confidence interval (95% CI), 1.5–3.4]. Individuals in the lowest tertile of OGG activity exhibited an increased risk of SCCHN with an OR of 7.0 (95% CI, 2.0–24.5). The combination of smoking and low OGG was associated with a highly increased estimated relative risk for SCCHN. These results suggest that low OGG is associated with the risk of SCCHN, and if confirmed by additional epidemiologic studies, screening of smokers for low OGG activity might be used as a strategy for the prevention of lung cancer and SCCHN. (Cancer Res 2006; 66(24): 11683-9)

Cancer is caused by mutations in critical genes that lead to uncontrolled cell proliferation (1). Therefore, the rate at which mutations accumulate plays a major role in cancer risk and is a potential target for cancer prevention. Two major factors that affect mutation rates are exposure to carcinogens, e.g., DNA-damaging agents such as sunlight or tobacco smoke, and the efficiency of cellular pathways that process DNA damage (2). Thus, the combination of exposure to a carcinogen and a reduced individual ability to remove DNA damage and restore the original DNA sequence might have a significant role in the etiology of human cancer (24). The importance of DNA repair in cancer pathogenesis became clear when it was discovered that several of the mutated genes responsible for hereditary cancer predisposition syndromes encode DNA-repair proteins. These include xeroderma pigmentosum, a disease characterized by extreme susceptibility to sunlight-associated skin cancer caused by a defect in nucleotide excision repair (5); hereditary nonpolyposis colorectal cancer, caused by mutations in mismatch repair genes (6, 7); and human MutY homologue (MutYH)–associated polyposis, caused by biallelic mutations in the base excision repair DNA glycosylase MYH (8).

Reduced DNA-repair capacity is also expected to play an important role in sporadic cancer. Indeed, it has been reported that single nucleotide polymorphisms in several DNA-repair genes (912), as well as suboptimal DNA-repair activity as assayed in peripheral blood mononuclear cells (PBMC; refs. 9, 1318), are associated with increased risk for a variety of epithelial cancers.

We have adopted a functional assay approach to studying the role of DNA repair in the etiology of human cancer and have recently shown that reduced repair of the oxidative DNA lesion 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxoguanine) is a risk factor for non–small cell lung cancer (NSCLC; ref. 19). This was accomplished using an enzymatic assay for 8-oxoguanine DNA glycosylase (OGG), the enzyme that removes 8-oxoguanine from DNA (2). The interest in 8-oxoguanine stems from the fact that it is a common mutagenic DNA lesion that is formed by intracellular oxidation as a byproduct of cellular metabolism as well as from exposure to external carcinogens, such as radiation and tobacco smoke (2022). Importantly, the combination of reduced OGG activity and smoking was associated with a high estimated relative risk of ∼120, illustrating the cumulative effect of genes and the environment and making the OGG test a potential screening tool for lung cancer prevention among smokers (19, 2325).

DNA repair is a housekeeping process that operates in essentially all cell types. This raises the question of tissue specificity in cancer risk caused by reduced OGG and its interrelationship with tobacco smoke. Here, we report that reduced OGG activity is associated with risk of squamous cell carcinoma of the head and neck (SCCHN), and that the combination of smoking and low OGG activity is associated with an even higher risk for the disease.

Study subjects. Head and neck cancer patients were enrolled from the Tel Aviv Sourasky Medical Center, and the cancer was histopathologically confirmed in all patients. Healthy control subjects were volunteers from the Sheba Medical Center and the Weizmann Institute of Science, including employees and retired employees and their relatives, frequency matched to the cases on age and gender. Each subject was interviewed by one of the team physicians. A standard questionnaire was used to collect information on demographic data and cancer risk factors, including smoking status. Exclusion criteria were prior chemotherapy or radiation therapy (for the case patients) or prior cancer (for control subjects). Out of the 49 head and neck cancer patients, 41 had SCCHN, and the other 8 patients had other histologies. The cancer sites were primarily the larynx (51%) and the oral cavity (16%). We analyzed the subgroup of 41 patients having SCCHN, although similar results were obtained with the entire group with 49 head and neck cancer patients. Blood samples were re-collected 3 to 4 years after diagnosis and treatment from 18 case patients who were defined as disease-free. This group consisted of individuals with a broad range of age, both males and females, and with early and advanced stages of SCCHN at diagnosis. All subjects provided written informed consent. The study was approved by the institutional ethics committees of the Sourasky Medical Center and the Weizmann Institute of Science.

The OGG activity test. To examine whether reduced OGG activity is associated with head and neck cancer risk, we used the OGG nicking assay (26), which was adapted in our laboratory to be of epidemiology grade (19). The assay, which measures the enzymatic removal of 8-oxoguanine from DNA, uses as a substrate a 32-bp synthetic DNA carrying a site-specific 8-oxoguanine, with a 32P radiolabel at the 5′ terminus of the DNA strand with the oxidized guanine. OGG activity is assayed by measuring the nicking of the radiolabeled strand, caused by the removal of the modified guanine by the OGG1 DNA glycosylase and the subsequent cleavage at the abasic site that it generated. The source of the OGG1 activity is a protein extract prepared from PBMC obtained from 10 mL of blood. Analysis by urea-PAGE followed by phosphorimaging is used to quantify the extent of nicking, indicated by the formation of the shorter radiolabeled DNA fragment. The assay is of epidemiology grade: It uses frozen PBMC and is relatively simple, reproducible (coefficient of variation is 10%), and stable for at least 3 years in the same individual (19).

Statistical methods. Cases and controls were classified into matched sets according to 5-year age groups and sex. Four cases and 27 controls were omitted from the analysis because they did not have a matching control or case. The remaining 37 cases and 93 controls are included in the analysis presented in this report, and the numbers in each matched set are shown in Table 1. All comparisons of cases with controls used conditional logistic regression, conditioning on the matched sets shown in Table 1. These analyses yielded odds ratio (OR) estimates. We have also done the statistical analysis with all 41 case patients and 41 control subjects matched according to 15-year age groups and sex and obtained very similar results. Statistical tests and confidence intervals (CI) for the ORs were based on the likelihood ratio test. Comparisons of OGG levels by stage of disease, age, sex, or smoking status within case or control groups used the Student's t test with equal or unequal variances as appropriate because OGG levels seemed to be distributed approximately normally. OGG values in 18 patients at diagnosis and after recovery following treatment were compared at the group level by a paired t test and at the individual level by calculating the CI values for the difference. These CI values were based on the within-person SD estimated from 23 replicated measures of an individual at the time of diagnosis and 28 replicated measures of an individual after recovery.

Table 1.

Age- and sex-matched sets for 93 controls and 37 head and neck cancer patients

Age (y)Male
Female
CasesControlsCasesControls
30–34 
40–44 
45–49 — — 
50–54 10 — — 
55–59 
60–64 
65–69 11 
70–74 12 
75–79 
80–84 
Total 26 49 11 44 
Age (y)Male
Female
CasesControlsCasesControls
30–34 
40–44 
45–49 — — 
50–54 10 — — 
55–59 
60–64 
65–69 11 
70–74 12 
75–79 
80–84 
Total 26 49 11 44 

NOTE: Analysis was performed also with 41 controls and 41 cases frequency matched according to 15-year age groups and sex.

OGG activity is reduced in PBMC from head and neck cancer patients. OGG activity was measured in 37 case patients with SCCHN and 93 control subjects frequency matched on age and sex. The mean OGG activity in PBMC from the SCCHN case patients was 6.1 units/μg of protein (95% CI, 5.6–6.5 units/μg of protein), statistically significantly lower than in controls subjects, which had a mean value of 7.2 units/μg of protein (95% CI, 7.0–7.4 units/μg of protein; P < 0.001; Table 2). Moreover, when we dichotomized the groups according to their age (<65 and ≥65 years old), sex (male and female), and smoking status (never-smoker and smoker), we found significantly lower mean OGG activity in all the subgroups of the case patients compared with control subjects, except for the subgroup of older (≥65 years old) individuals, where the lower OGG activity was not statistically significantly lower than in the control subjects (Table 2).

Table 2.

Distribution of selected characteristics and DNA-repair OGG activity values between head and neck cancer case patients and control subjects

VariableControl subjects (N = 93)
Case patients (N = 37)
P*
No.OGG activity (95% CI)No.OGG activity (95% CI)
All 93 7.2 (7.0–7.4) 37 6.1 (5.6–6.5) <0.001 
Age (y)      
    <65 50 7.5 (7.2–7.8) 16 5.7 (5.0–6.4) 0.001 
    ≥65 43 6.8 (6.6–7.1) 21 6.4 (5.8–7.0) 0.19 
  P = 0.001  P = 0.11  
Sex      
    Male 49 7.3 (7.0–7.6) 26 6.3 (5.8–6.8) 0.009 
    Female 44 7.1 (6.8–7.4) 11 5.5 (4.6–6.4) 0.01 
  P = 0.3  P = 0.09  
Smoking status      
    Never-smoker 70 7.1 (6.9–7.4) 18 5.9 (5.2–6.6) 0.03 
    Smoker 23 7.3 (7.0–7.6) 16 6.2 (5.5–6.9) 0.03 
  P = 0.47  P = 0.60  
Severity      
    Low (stages I and II)   15 6.7 (6.3–7.1)  
    High (stages III and IV)   22 5.7 (5.0–6.3)  
    P = 0.007  
VariableControl subjects (N = 93)
Case patients (N = 37)
P*
No.OGG activity (95% CI)No.OGG activity (95% CI)
All 93 7.2 (7.0–7.4) 37 6.1 (5.6–6.5) <0.001 
Age (y)      
    <65 50 7.5 (7.2–7.8) 16 5.7 (5.0–6.4) 0.001 
    ≥65 43 6.8 (6.6–7.1) 21 6.4 (5.8–7.0) 0.19 
  P = 0.001  P = 0.11  
Sex      
    Male 49 7.3 (7.0–7.6) 26 6.3 (5.8–6.8) 0.009 
    Female 44 7.1 (6.8–7.4) 11 5.5 (4.6–6.4) 0.01 
  P = 0.3  P = 0.09  
Smoking status      
    Never-smoker 70 7.1 (6.9–7.4) 18 5.9 (5.2–6.6) 0.03 
    Smoker 23 7.3 (7.0–7.6) 16 6.2 (5.5–6.9) 0.03 
  P = 0.47  P = 0.60  
Severity      
    Low (stages I and II)   15 6.7 (6.3–7.1)  
    High (stages III and IV)   22 5.7 (5.0–6.3)  
    P = 0.007  

NOTE: OGG activity was measured as described in Materials and Methods. OGG values are given as mean (95% CI), in enzyme units per microgram of protein extract. A similar analysis that was performed with 41 controls and 41 cases frequency matched according to 15-year age groups and sex yielded similar results.

*

Results of comparison using conditional logistic regression in the matched sets shown in Table 1.

Results of two-sided Student's t tests with equal or unequal variance as appropriate.

We next compared the distribution of OGG activity values in head and neck case patients with control subjects. Overall, the distribution of OGG activity values in case patients was shifted to lower values than in control subjects, with 19% of case patients but only 4% of control subjects having OGG activity values ≤5.5 (selected as an arbitrary cutpoint for low OGG activity; Fig. 1).

Figure 1.

Distribution of the DNA-repair activity of OGG in PBMC from patients with SCCHN and control subjects. OGG enzymatic activity was measured in cell extracts and quantified as enzyme units per microgram of protein extract. OGG activity of 93 control subjects (top) and 37 case patients (bottom).

Figure 1.

Distribution of the DNA-repair activity of OGG in PBMC from patients with SCCHN and control subjects. OGG enzymatic activity was measured in cell extracts and quantified as enzyme units per microgram of protein extract. OGG activity of 93 control subjects (top) and 37 case patients (bottom).

Close modal

When case subjects were analyzed according to stage at the time of diagnosis, we found a significantly lower mean OGG activity of 5.7 units/μg of protein (95% CI, 5.0–6.3 units/μg of protein) in patients with advanced stages of the disease (stages III and IV) compared with patients with early stages of the disease (stages I and II; mean OGG activity was 6.7 units/μg of protein; 95% CI, 6.3–7.1 units/μg of protein), with P = 0.007 (Table 2). Although this difference may be explained by a more aggressive SCCHN caused by lower OGG activity, it may also mean that SCCHN affects OGG activity. If the latter holds, it would be impossible to determine whether reduced OGG is a risk factor in SCCHN using the design of a case-control study.

OGG activity in case subjects is similar at diagnosis and after recovery from the disease. To examine whether the disease affected OGG activity in PBMC, we reassayed OGG activity in some case patients who were disease-free 3 to 4 years after therapy. As can be seen in Table 3, the retested group was representative of the entire case group in terms of age, gender, and stage. Fig. 2 shows the OGG activity in these 18 case patients at diagnosis and after recovery. All 18 disease-free individuals had OGG activity values within the 95% CI of their baseline OGG activity (Fig. 2A). In addition, the group mean value did not seem to change from diagnosis to recovery (P = 0.34 using the paired t test analysis). Thus, the OGG activity of patients who were successfully treated for head and neck cancer is similar to their OGG activity measured at diagnosis, suggesting no effect of SCCHN on OGG activity.

Table 3.

Characterization of head and neck cancer patients who were assayed for OGG activity at diagnosis and after recovery

No.SexAgeSmokingDisease stageOGG activity at diagnosisOGG activity after recovery
49 Yes II 6.9 7.1 
63 Yes II 7.3 6.3 
67 Yes 8.3 9.6 
4* 56 No II 7.6 7.6 
76 No 7.1 6.3 
64 Yes IV 3.1 4.2 
80 Yes II 6.0 6.2 
8* 59 No II 6.0 6.4 
82 No II 5.6 6.7 
10 72 Yes 6.0 6.7 
11 54 Yes IV 7.0 5.7 
12 88 No IV 5.0 5.6 
13 70 No IV 5.6 5.7 
14 56 Yes 6.5 5.8 
15 73 No 7.2 7.7 
16 63 Yes IV 5.8 5.1 
17 59 Yes III 5.8 6.6 
18 42 Yes IV 6.0 7.3 
No.SexAgeSmokingDisease stageOGG activity at diagnosisOGG activity after recovery
49 Yes II 6.9 7.1 
63 Yes II 7.3 6.3 
67 Yes 8.3 9.6 
4* 56 No II 7.6 7.6 
76 No 7.1 6.3 
64 Yes IV 3.1 4.2 
80 Yes II 6.0 6.2 
8* 59 No II 6.0 6.4 
82 No II 5.6 6.7 
10 72 Yes 6.0 6.7 
11 54 Yes IV 7.0 5.7 
12 88 No IV 5.0 5.6 
13 70 No IV 5.6 5.7 
14 56 Yes 6.5 5.8 
15 73 No 7.2 7.7 
16 63 Yes IV 5.8 5.1 
17 59 Yes III 5.8 6.6 
18 42 Yes IV 6.0 7.3 
*

These patients were excluded from the case-control study due to non-SCC histology (patient 4) and prior chemotherapy (patient 8).

Figure 2.

Comparison between OGG activities at diagnosis and after recovery in SCCHN patients. OGG activities of 18 patients were measured at diagnosis and after recovery from the disease (3–4 years after diagnosis and treatment; with no evidence for disease). A, OGG values with the 95% CI at diagnosis (□) and after recovery (▪). B, differences in OGG values as measured at diagnosis (OGG1) and after recovery (OGG2) for each individual (▪). Vertical lines, 95% CI of the differences calculated based on the 10% coefficient of variation of the OGG assay.

Figure 2.

Comparison between OGG activities at diagnosis and after recovery in SCCHN patients. OGG activities of 18 patients were measured at diagnosis and after recovery from the disease (3–4 years after diagnosis and treatment; with no evidence for disease). A, OGG values with the 95% CI at diagnosis (□) and after recovery (▪). B, differences in OGG values as measured at diagnosis (OGG1) and after recovery (OGG2) for each individual (▪). Vertical lines, 95% CI of the differences calculated based on the 10% coefficient of variation of the OGG assay.

Close modal

Reduced OGG activity is associated with the risk of SCCHN. To determine whether there was an association between reduced OGG activity and risk of head and neck cancer, we used conditional logistic regression, in which the presence or absence of SCCHN was a binary-dependent variable; age was a continuous variable; smoking status (smoker versus never-smokers) was a dichotomous variable; and OGG activity was expressed as a continuous measure, in tertiles or as a binary variable (above or below the control median). Reduced OGG activity was found to be associated with SCCHN, as indicated by the statistically significant ORs obtained using all three conditional logistic regression models. When OGG activity was used as a continuous variable, with adjustment for age and smoking status, the adjusted OR for head and neck cancer associated with a 1-unit decrease in OGG activity was 2.3 (95% CI, 1.5–3.4; P < 0.0001; Table 4). When OGG activity was dichotomized, the adjusted OR for SCCHN associated with low OGG activity was 7.3 (95% CI, 2.5–21.1; P = 0.0003; Table 4). In the tertile model, using the tertile with the highest OGG activity as the reference group, the adjusted ORs for SCCHN were 3.7 (95% CI, 1.3–10.6; P = 0.02) for the middle tertile and 7.0 (95% CI, 2.0–24.5; P = 0.002) for the lowest tertile (Table 4).

Table 4.

Conditional logistic regression analysis of OGG activity value in head and neck cancer patients and control subjects

VariableCase subjects, n (%)Control subjects, n (%)Adjusted* OR (95% CI)
OGG 37 (100.0) 93 (100.0) 2.3 (1.5–3.4) 
   P < 0.0001 
OGG (median)    
    >7.3 4 (10.8) 46 (49.5) 1.0 (reference) 
    ≤7.3 33 (89.2) 47 (50.5) 7.3 (2.5–21.1) 
   P = 0.0003 
OGG (tertiles)    
    >7.6 3 (8.1) 31 (33.3) 1.0 (reference) 
    6.8–7.6 8 (21.6) 31 (33.3) 3.7 (1.3–10.6) 
   P = 0.02 
    <6.8 26 (70.3) 31 (33.3) 7.0 (2.0–24.5) 
   P = 0.002 
    Trend test   P = 0.0007 
VariableCase subjects, n (%)Control subjects, n (%)Adjusted* OR (95% CI)
OGG 37 (100.0) 93 (100.0) 2.3 (1.5–3.4) 
   P < 0.0001 
OGG (median)    
    >7.3 4 (10.8) 46 (49.5) 1.0 (reference) 
    ≤7.3 33 (89.2) 47 (50.5) 7.3 (2.5–21.1) 
   P = 0.0003 
OGG (tertiles)    
    >7.6 3 (8.1) 31 (33.3) 1.0 (reference) 
    6.8–7.6 8 (21.6) 31 (33.3) 3.7 (1.3–10.6) 
   P = 0.02 
    <6.8 26 (70.3) 31 (33.3) 7.0 (2.0–24.5) 
   P = 0.002 
    Trend test   P = 0.0007 

NOTE: OGG activity was measured as described in Materials and Methods and was fitted in the conditional logistic regression model as a continuous variable. There was no interaction between OGG and smoking (P = 0.40). A similar analysis that was performed with 41 controls and 41 cases frequency matched according to 15-year age groups and sex yielded similar results.

*

Conditional logistic regression for matched sets adjusted for smoking status (smoker, nonsmoker). P and CI values were based on the likelihood ratio test.

Combined effect of reduced OGG activity and smoking on the estimated relative risk for SCCHN. As in the case of lung cancer, we found no statistical interaction between smoking and reduced OGG activity (P = 0.40), implying that the two are independent risk factors for SCCHN. We used the conditional logistic regression model with OGG activity as a continuous variable to determine the association of smoking with SCCHN. We found that smoking was associated with head and neck cancer, with an OR of 5.8 (95% CI, 1.8–18.4; P = 0.001). This value is similar to the result of other studies showing that smoking is associated with SCCHN (27).

Using conditional logistic regression, we estimated the relative risk of SCCHN for the combination of smoking and reduced OGG activity. As can be seen in Fig. 3, the estimated risk for SCCHN was 13 (95% CI, 3.7–46.1), 30 (95% CI, 6.9–128.3), and 68 (95% CI, 12–383) times higher for smokers who had OGG lower by 1, 2, or 3 units than for nonsmokers with a normal OGG activity of 7.2 units/μg of protein.

Figure 3.

The estimated relative risk for SCCHN associated with the combination of low OGG and smoking. Results of logistic regression for the indicated decrease in OGG activity either with (black columns) or without smoking (gray columns). See text for details.

Figure 3.

The estimated relative risk for SCCHN associated with the combination of low OGG and smoking. Results of logistic regression for the indicated decrease in OGG activity either with (black columns) or without smoking (gray columns). See text for details.

Close modal

DNA repair is a major mechanism for maintaining genome integrity and preventing mutations. Given the critical role of mutation in carcinogenesis, it was hypothesized long ago that suboptimal DNA repair might be an important factor in the etiology of human cancer (2). This hypothesis was clearly proven for hereditary cancers, where mutations in DNA-repair genes were shown to cause cancer predisposition syndromes as discussed above. The role of DNA repair in the etiology of sporadic cancer is likely to be similarly important but is more difficult to prove. One of the reasons is that unlike in hereditary cancer, where a big deficiency in the activity of a single gene is the main cause of the disease, sporadic cancer is caused most likely by subtle changes in the activity of multiple genes, combined with environmental and lifestyle factors (9, 15). This means that generally small individual effects of DNA-repair activities need to be assayed, which puts a bigger burden on the quality of the assays used to evaluate the role of DNA repair in sporadic cancer. Nevertheless, there is a growing body of literature on the importance of DNA repair in sporadic cancer in general (reviewed in refs. 9, 15) and head and neck cancer in particular (2730). Because DNA repair consists of multiple pathways that operate on different (but partially overlapping) types of DNA damage, we adopted the approach of developing specific functional DNA-repair assays, with the expectation that it will provide data with higher resolution on the involvement of specific repair activities in cancer risk. Indeed, using the OGG assay, which specifically measures the repair of 8-oxoguanine–containing DNA, we show here that reduced OGG activity is associated with an increased risk of SCCHN. The small 15% difference in the mean OGG activity between healthy individuals (7.2 units/μg of protein) and patients with SCCHN (6.1 units/μg of protein) is somewhat misleading because it averages OGG activity of case patients in which low OGG is a risk factor, with case patients who have normal OGG activity and therefore are not affected by this risk factor. This is clearly illustrated in Table 4 and Fig. 3, which show that for individuals who have OGG activity lower by 2 or 3 OGG activity units per microgram of protein than the mean (reductions of 28% or 42% from the mean, respectively), the estimated relative risk (approximated by the OR) is 5.3 and 12.2, respectively.

Case-control studies like the one reported here have two major limitations. (a) The assay is done on a surrogate tissue, PBMC. In our previous study on the association between OGG activity and the risk of NSCLC, we have shown that the OGG activity in PBMC correlates with the OGG activity in nontumor lung tissue from patients, justifying the use of PBMC as a surrogate to the lung (19). Such a measurement was not possible in this study; however, the lung cancer study suggests that PBMC might provide a reasonable surrogate for comparing OGG activity between individuals, despite the fact that within an individual, there might be a tissue-specific expression of OGG activity. (b) Because of the retrospective study design, it is not clear whether the reduced activity observed in case patients was indeed a cause or a result of the disease. To address this question, we retested patients after treatment and follow-up. Our results clearly show that OGG activities were very similar at diagnosis and after recovery. Although this result does not absolutely prove that low OGG activity is a risk factor, it does support this interpretation.

Tobacco smoking is an important determinant of SCCHN risk (reviewed in ref. 27), as we have also shown in our analysis (OR, 5.8; 95% CI, 1.8–18.4). We found that smoking had no effect on OGG activity in PBMC, which differs from an earlier report in which OGG activity in PBMC of smokers was found to be higher than nonsmokers (20). We do not know the reasons for the different results, but note that our study included a much larger group of individuals of 130, compared with only 30 in the other study (20). We also found no interaction between OGG and smoking. The combination of reduced OGG and smoking was associated with a markedly increased risk for SCCHN, reaching OR values of 60 and even higher, for OGG values below 4 units/μg of protein. The simplest explanation for these results is the following. Tobacco smoke causes a load of DNA damage, including 8-oxoguanine. Reduced DNA repair is unable to cope with the extra number of lesions, leading to an increase in the number of mutations and, therefore, to increased cancer risk. A similar cumulative effect of reduced OGG and smoking was previously found for NSCLC (19). Larger epidemiologic studies will enable the classification of smoking by dose (13), thereby providing better insight into the combined effects of smoking and low OGG activity.

Interestingly, OGG activity was lower in patients with stages III and IV of SCCHN compared with patients with stages I and II (Table 2). This finding can have at least two explanations: (a) a bigger effect of the tumor in its advanced stages on OGG activity in PBMC and (b) faster disease progression when OGG activity is lower. Because the presence of the tumor does not seem to affect OGG activity (Fig. 2; Table 3), a low OGG activity might be associated with a more aggressive disease. However, the relatively small sample size of this study prohibits a definitive conclusion, and larger epidemiologic studies are needed to confirm this possibility.

The OGG1 gene contains the Ser326Cys polymorphism, which is common in the population (31). Most, but not all, studies report that this polymorphism is associated with reduced enzymatic activity (3235). In addition, an altered intracellular localization behavior of the variant OGG1 protein was reported (36). Epidemiologic studies on the association between the OGG1 Ser326Cys polymorphism and head and neck cancer have yielded conflicting results (37, 38). Notably, the larger study, which was done with 706 cases and 1,196 controls, showed no association of this polymorphism with SCCHN (38). It should be emphasized that although functional DNA-repair assays are more laborious compared with genotyping, generally, they are likely to characterize the DNA-repair capacity of an individual more accurately than a genetic polymorphism because (a) a genetic polymorphism is not always correlated with function, and even if it is, its effect may be very small or compensated by other factors; (b) the activity of DNA-repair proteins and enzymes, like the activity of any other proteins, might be affected by numerous factors, including the level of transcription, splicing, the stability of the mRNA, translation, protein stability, posttranslational modification, and the action of inhibitors or stimulators. In addition, there may be epigenetic factors (e.g., methylation; refs. 39, 40) or stochastic factors that affect the enzyme activity. A functional assay is a sum of a large number of these factors and may be equivalent to the sum of dozens, if not hundreds, of polymorphisms plus at least some environmental and lifestyle factors. Thus, although polymorphisms may be useful in assessing risk in some cases, in many cases where multiple factors affect activity, a functional activity assay may be advantageous (25).

DNA repair is a fundamental housekeeping cellular operation and is therefore expected to affect all cell types. Nevertheless, hereditary deficiencies in DNA-repair activities exhibit a predisposition to specific types of cancer, as described above, for reasons that are only partially understood. Reduced OGG activity could be expected to be a risk factor in other smoking-related cancers. However, given the abundance of 8-oxoguanine and the suspected role of oxidative stress in cancer, reduced OGG activity might be associated with the risk of some other cancers as well.

The study presented here is a pilot case-control study, which yielded statistically significant results, because there was a strong association between reduced OGG and SCCHN. The large, estimated relative risk for SCCHN of smoking and low OGG combined suggest that screening of smokers for low OGG activity may, in the future, provide a persuasive tool for intensive smoking cessation intervention or enrollment in early detection programs. Such an approach, which may be extended to include additional DNA-repair assays, may provide an effective strategy for the prevention of tobacco-related cancers. However, larger case-control studies are needed to validate these results, and prospective epidemiologic studies are needed to prove that reduced OGG is indeed a risk factor for SCCHN.

Note: Z. Livneh is the incumbent of The Maxwell Ellis Professorial Chair in Biomedical Research.

Grant support: Flight Attendants Medical Research Institute, Florida, National Cancer Institute, NIH grant U01 CA111219, and the M.D. Moross Institute for Cancer Research at the Weizmann Institute of Science (Z. Livneh).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Margaret R. Spitz (Houston, TX) and Gad Rennert (Haifa, Israel) for helpful epidemiologic advice.

1
Vogelstein B, Kinzler KW. Cancer genes and the pathways they control.
Nat Med
2004
;
10
:
789
–99.
2
Friedberg EC, Walker GC, Siede W, Wood RD, Schultz RA, Ellenberger T. DNA repair and mutagenesis. 2nd ed. Washington (DC): ASM Press; 2006.
3
Vogelstein B, Kinzler KW. The genetic basis of human cancer. New York: McGraw-Hill; 1998.
4
Wood RD, Mitchell M, Sgouros J, Lindahl T. Human DNA repair genes.
Science
2001
;
291
:
1284
–9.
5
Weeda G, Hoeijmakers JHJ, Bootsma D. Genes controlling nucleotide excision repair in eukaryotic cells.
Bioessays
1993
;
15
:
249
–58.
6
Modrich P. Mismatch repair, genetic stability, and cancer.
Science
1994
;
266
:
1959
–60.
7
Kolodner RD. Mismatch repair: mechanisms and relationship to cancer susceptibility.
Trends Biochem Sci
1995
;
20
:
397
–401.
8
Sieber OM, Lipton L, Crabtree M, et al. Multiple colorectal adenomas, classic adenomatous polyposis, and germ-line mutations in MYH.
N Engl J Med
2003
;
348
:
791
–9.
9
Berwick M, Vineis P. Markers of DNA repair and susceptibility to cancer in humans: an epidemiologic review.
J Natl Cancer Inst
2000
;
92
:
874
–97.
10
Goode EL, Ulrich CM, Potter JD. Polymorphisms in DNA repair genes and associations with cancer risk.
Cancer Epidemiol Biomarkers Prev
2002
;
11
:
1513
–30.
11
Hung RJ, Hall J, Brennan P, Boffetta P. Genetic polymorphisms in the base excision repair pathway and cancer risk: a HuGE review.
Am J Epidemiol
2005
;
162
:
925
–42.
12
Sakiyama T, Kohno T, Mimaki S, et al. Association of amino acid substitution polymorphism in DNA repair genes TP53, POLI, REV1 and LIG4 with lung cancer risk.
Int J Cancer
2005
;
114
:
730
–7.
13
Wei Q, Cheng L, Amos CI, et al. Repair of tobacco carcinogen-induced DNA adducts and lung cancer risk: a molecular epidemiologic study.
J Natl Cancer Inst
2000
;
92
:
1764
–72.
14
Rajaee-Behbahani N, Schmezer P, Risch A, et al. Altered DNA repair capacity and bleomycin sensitivity as risk markers for non-small cell lung cancer.
Int J Cancer (Pred Oncol)
2001
;
95
:
86
–91.
15
Spitz MR, Wei Q, Dong Q, Amos CI, Wu X. Genetic susceptibility to lung cancer: the role of DNA damage and repair.
Cancer Epidemiol Biomarkers Prev
2003
;
12
:
689
–98.
16
Speina E, Zielinska M, Barbin A, et al. Decreased repair activities of 1,N(6)-ethenoadenine and 3,N(4)-ethenocytosine in lung adenocarcinoma patients.
Cancer Res
2003
;
63
:
4351
–7.
17
Gackowski D, Speina E, Zielinska M, et al. Products of oxidative DNA damage and repair as possible biomarkers of susceptibility to lung cancer.
Cancer Res
2003
;
63
:
4899
–902.
18
Cheng L, Eicher SA, Guo Z, Hong WK, Spitz MR, Wei Q. Reduced DNA repair capacity in head and neck cancer patients.
Cancer Epidemiol Biomarkers Prev
1998
;
7
:
465
–8.
19
Paz-Elizur T, Krupsky M, Blumenstein S, Elinger D, Schechtman E, Livneh Z. Reduced DNA repair activity for oxidative damage and the risk of lung cancer.
J Natl Cancer Inst
2003
;
95
:
1312
–9.
20
Asami S, Hirano T, Yamaguchi R, Tomioka Y, Itoh H, Kasai H. Increase of a type of oxidative DNA damage, 8-hydroxyguanine, and its repair activity in human leukocytes by cigarette smoking.
Cancer Res
1996
;
56
:
2546
–9.
21
Cooke MS, Evans MD, Dizdaroglu M, Lunec J. Oxidative DNA damage: mechanisms, mutation, and disease.
FASEB J
2003
;
17
:
1195
–214.
22
Wallace SS. Biological consequences of free radical-damaged DNA bases.
Free Radic Biol Med
2002
;
33
:
1
–14.
23
Caporaso N. The molecular epidemiology of oxidative damage to DNA and cancer.
J Natl Cancer Inst
2003
;
95
:
1263
–5.
24
Paz-Elizur T, Krupsky M, Elinger D, Schechtman E, Livneh Z. Repair of the oxidative DNA damage 8-oxoguanine as a biomarker for lung cancer risk.
Cancer Biomarkers
2005
;
1
:
201
–5.
25
Paz-Elizur T, Brenner DE, Livneh Z. Interrogating DNA repair in cancer risk assessment.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
1585
–7.
26
Chung MH, Kasai H, Jones DS, et al. An endonuclease activity of Escherichia coli that specifically removes 8-hydroxyguanine residues from DNA.
Mutat Res
1991
;
254
:
1
–12.
27
Sturgis EM, Wei Q, Spitz MR. Descriptive epidemiology and risk factors for head and neck cancer.
Semin Oncol
2004
;
31
:
726
–33.
28
Hung RJ, van der Hel O, Tavtigian SV, Brennan P, Boffetta P, Hashibe M. Perspectives on the molecular epidemiology of aerodigestive tract cancers.
Mutat Res
2005
;
592
:
102
–18.
29
Wei Q, Wang L-E, Sturgis EM, Mao L. Expression of nucleotide excision repair proteins in lymphocytes as a marker of susceptibility to squamous cell carcinima of the head and neck.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
1961
–6.
30
Licitra L, Rossini C, Bossi P, Locati LD. Advances in the changing patterns of aetiology of head and neck cancers.
Curr Opin Otolaryngol Head Neck Surg
2006
;
14
:
95
–9.
31
Sugimura H, Kohno T, Wakai K, et al. hOGG1 Ser326Cys polymorphism and lung cancer susceptibility.
Cancer Epidemiol Biomarkers Prev
1999
;
8
:
669
–74.
32
Dherin C, Radicella JP, Dizdaroglu M, Boiteux S. Excision of oxidatively damaged DNA bases by the human α-hOGG1 protein and the polymorphic α-hOGG1-(Ser326/Cys326) protein which is frequently found in human populations.
Nucleic Acids Res
1999
;
27
:
4001
–7.
33
Janssen K, Schlink K, Gotte W, Hippler B, Kaina B, Oesch F. DNA repair activity of 8-oxoguanine DNA glycosylase I (OGG1) in human lymphocytes is not dependent on genetic polymorphism Ser326/Cys326.
Mutat Res
2001
;
486
:
207
–16.
34
Yamane A, Kohno T, Ito K, et al. Differential ability of polymorphic OGG1 proteins to suppress mutagenesis induced in human cell in vivo.
Carcinogenesis
2004
;
25
:
1689
–94.
35
Hill JW, Evans MK. Dimerization and opposite base-dependent catalytic impairment of polymorphic S326C OGG1 glycosylase.
Nucleic Acids Res
2006
;
34
:
1620
–32.
36
Luna L, Rolseth V, Hilrestrand GA, et al. Dynamic relocalization of hOGG1 during the cell cycle is disrupted in cells harboring the hOGG1-326 polymorphic variant.
Nucleic Acids Res
2005
;
33
:
1813
–24.
37
Elahi A, Zheng Z, Park J, Eyring K, McCaffrey T, Lazarus P. The human OGG1 DNA repair enzyme and its association with orolaryngeal cancer risk.
Carcinogenesis
2002
;
23
:
1229
–34.
38
Zhang Z, Shi Q, Wang LE, et al. No association between hOGG1 Ser326Cys polymorphism and risk of squamous cell carcinoma of the head and neck.
Cancer Epidemiol Biomarkers Prev
2004
;
13
:
1081
–3.
39
Merlo A, Herman JG, Mao L, et al. 5′ CpG island methylation is associated with transcriptional silencing of the tumor suppressor p16/CDKN2/MTS1 in human cancers.
Nat Med
1995
;
1
:
686
–92.
40
Esteller M, Herman JG. Cancer as an epigenetic disease: DNA methylation and chromatin alterations in human tumours.
J Pathol
2002
;
196
:
1
–7.