September 19, 2003 Access past issues

Reviewed by: Shauna Lyn, MD Marta Gwinn, MD, MS Muin J. Khoury, M.D. PhD Office of Genomics and Disease Prevention Centers for Disease Control and Prevention

The Health Outcome

Lung cancer is the most common form of cancer in the world today, accounting for 12.3% of all cancers and 17.8% of cancer- related deaths worldwide in 2000.  Tobacco smoking has long been known to be the primary risk factor for this cancer. There is a known dose- response relationship between lung cancer risk and the number of cigarettes smoked per day, the degree of inhalation, and the age at initiation; a lifetime smoker has a lung cancer risk 20-30 times that of a non- smoker (1). Cessation of smoking results in a decreased risk of lung cancer after a lag period of about seven years, but the risk never reaches baseline levels (2).

Over 20 lung- cancer- specific carcinogens in tobacco are known to chemically modify DNA, leading to genetic mutations. Accumulation of these mutations in critical oncogenes and tumor suppressor genes promotes cancer development.  A person with low DNA repair capacity is less able to counteract these carcinogenic effects, and is therefore at increased risk for lung cancer (3,4) .  Still, although more than 80% of lung cancers are attributed to tobacco, only 15% of smokers ultimately develop lung cancer (5).  This, as well as the fact that a family history of lung cancer increases its risk (irrespective of smoking history), (1,2)  suggests that genetic factors may also predispose one to the risk of lung cancer.  Polymorphisms of a number of genes involved in the metabolism of carcinogens and the repair of DNA have been shown to influence the development of lung cancer (6-8).

Reactive oxygen species (ROS) formed as by- products of normal cellular metabolism and generated by oxidative stresses from the environment (eg, ionizing radiation or chemical genotoxic compounds) damage DNA and are thought to play a key role in mutagenesis and carcinogenesis (9).  One such DNA lesion, 8- oxyguanine (8-OG), is highly mutagenic, causing GC to TA transversions. It is induced in lung DNA both by ROS and by cigarette smoking (10). Elevated amounts of 8-OG have also been found in peripheral leukocytes and lung tissue of smokers and lung cancer patients (11).  The human 8- oxoguanine DNA glycosylase gene (OGG1), is localized on chromosome 3p26 and encodes a DNA glycosylase (OGG) that excises 8-OG from damaged DNA (12).

The Finding

Paz- Elizur, et al (#8) reported that low OGG activity is associated with an increased risk of lung cancer. This conclusion was drawn from a case- control study conducted from April 1999 through January 2002 in Israel . The study included 68 case patients with operable non- small- cell lung cancer (NSCLC) and 68 healthy control subjects, frequency matched for age and sex. Blood samples were taken from participants at different time points, and protein extracts were made from isolated peripheral blood mononuclear cells (PBMCs). Protein extracts were also made from non- tumor lung tissue of case patients who underwent primary lobectomy or pneumonectomy.

The authors first developed an assay for the ability of OGG to remove 8-OG from DNA. A synthesized oligonucleotide containing a site- specific 8-OG residue was incubated with protein extracts from each participant’s PBMCs and lung tissue, which contained their OGG. Specific OGG activity was calculated by dividing the measured enzyme activity by the amount of protein in the reaction mixture. In assays of lung tissue, OGG was normalized to the total amount of DNA rather than the total amount of protein in an attempt to better represent the intracellular contents.

In the case- control study, the authors found that the mean OGG activity was significantly lower in case subjects than in controls (p<0.001); they found no significant differences in OGG activity between men and women or between smokers and non- smokers. They performed a conditional logistic regression analysis of the relationship between OGG and lung cancer, adjusting for age and smoking status. OGG activity was analyzed as a continuous variable and also using two different categorical variables (defined either by the median value in controls or by tertiles). All three models found a statistically significant effect for OGG activity. The adjusted odds ratio for lung cancer comparing the lowest tertile (OGG activity < 6.7 U/ μg protein) with the highest tertile (>7.5 U/ μg protein) was 4.8 (95% CI = 1.5 to 15.9, p= 0.01).  Smoking was strongly associated with NSCLC with an odds ratio of 18 (95% CI = 6.0 to 53, p<0.001) when controlling for OGG activity as a continuous variable. The authors’ estimated relative risks of lung cancer for smokers can be summarized in a 2 by 4 table of ORs based on this model: 

Smoking status                                      OGG activity   7.0 U/ μg 6.0 U/ μg 4.0 U/ μg         Nonsmoker 1.0 1.9 7.0 Smoker 18 34 124

These findings are consistent with no interaction according to a multiplicative model, suggesting that they operate independently as risk factors for lung cancer. The authors draw this conclusion, without presenting a statistical test for interaction.

The authors propose that screening smokers for low OGG activity and targeting smoking cessation interventions to those with increased genetic susceptibility may lead to a decrease in the incidence of lung cancer. They also propose that nonsmokers with low OGG activity (who are also at elevated risk of lung cancer, though not as high as smokers), should avoid exposures that can produce DNA damage (eg, ionizing radiation and secondary smoke). They state that prospective epidemiologic studies are first required to confirm their findings.

Public Health Implications

The relationship between OGG1 and cancer risk has been of interest since the gene was first cloned in 1997. The results of studies hoping to demonstrate the contribution of OGG1 polymorphisms to lung cancer risk in smokers have been somewhat inconsistent. For example, two studies in Japan found a significant association between Ser326Cys and squamous cell carcinoma but not other types of lung cancer (13,14), and two other studies (11,15), reported ORs ≤ 2 for the association between Ser326Cys and lung cancer. The authors of this study point out that OGG1 polymorphisms are only one of many factors (including natural inhibitors and activators, posttranslational modifications, differences in levels of expression and stability) that might affect OGG activity, and that,  “…because activity measurements are directly associated with protein function, they should be more effective than gene- specific polymorphisms as predictors of cancer risk.”

This study by Paz-Elizur, et al offers some preliminary evidence that OGG activity may predict patients’ risk of lung cancer. However, factors such as selection bias and issues concerning the validity of the assay they developed raise more questions that have to be answered before this OGG activity assay can be used to predict lung cancer risk (see accompanying abstract for detailed comments). More steps must be taken to validate this assay before its utility in epidemiologic studies of lung cancer can be determined. A prospective study (as suggested by the authors) to evaluate the predictive value of OGG activity may not be very feasible, given that lung cancer occurs infrequently, even in smokers, and takes decades to develop. A case- control study nested within an existing cohort would be a better choice. Finally, the suggestion by the authors that smoking cessation measures be targeted at individuals who have the advantage of knowing they are genetically susceptible to lung cancer misses the mark. Given the great detriment to health that tobacco has proven itself to be over the years, there should be no limited recommendations with respect to who should not smoke. If the goal is optimal health for all people, no one should smoke.

References

1. Parkin DM, et al. Case burden in the year 2000. The global picture. European Journal of Cancer. 2001; 37: S4-S66.
2. Minna J, et al. Focus on lung cancer. Cancer Cell. 2002; 1: 49 -52.
3. Berwick M, et al. Markers of DNA repair and susceptibility to cancer in humans: an epidemiologic review. 2000; 92: 874- 897.
4. Wei Q, et al. Repair of tobacco carcinogen- induced DNA adducts and lung cancer risk: a molecular epidemiologic study. Journal of the National Cancer Institute. 2000; 92: 1764- 1772.
5. Spitz MR, et al. Genetic susceptibility to lung cancer: the role of DNA damage and repair. Cancer Epidemiology, Biomarkers & Prevention. 2003; 12: 689- 698.
6. Bartsch, H, et al. Genetic polymorphism of CYP genes, alone or in combination, as a risk modifier of tobacco- related cancers. Cancer Epidemiology, Biomarkers & Prevention. 2000; 9: 3-28.
7. Zhou w, et al. Polymorphisms in the DNA repair genes XRCC1 and ERCC2, smoking, and lung cancer risk. Cancer Epidemiology, Biomarkers & Prevention, 2003; 12: 359-365.
8. Paz-Elizur T, et al. DNA repair activity for oxidative damage and risk of lung cancer. Journal of the National Cancer Institute. 2003; 95: 1312- 1319.
9. Boiteux S, et al. The human OGG1 gene: structure, functions, and its implication in the process of carcinogenesis. Archives of Biochemistry and Biophysics; 377: 1-8.
10. Asami S, et al. Cigarette smoking induces an increase in oxidative DNA damage, 8- hydroxydeoxyguanosine, in a central site of the human lung. Carcinogenesis. 1997; 18: 1763- 1766.
11. Le Marchand L, et al. Association of the hOGG1 Ser326Cys polymorphism with lung cancer risk. Cancer Epidemiology, Biomarkers & Prevention. 2002; 11: 409- 412.
12. McKusick VA , et al. 8- oxoguanine DNA glycosylase; OGG1 Accessed September 19, 2003 Website.
13. Sugimara H, et al. hOGG1 Ser326Cys polymorphism and lung cancer susceptibility. Cancer Epidemiology, Biomarkers & Prevention. 1999; 8: 669-674.
14. Ito H, et al. A limited association of OGG1 Ser326Cys polymorphism for adenocarcinoma of the lung. Journal of Epidemiology. 2002; 12: 258- 265.
15. Wikman H, et al. hOGG1 polymorphism and loss of heterozygosity (LOH): significance for lung cancer susceptibility in a Caucasian population. International Journal of Cancer. 2000; 88: 932- 937.