Published ahead of print on September 1, 2005, doi:10.1164/rccm.200502-235PP
© 2006 American Thoracic Society doi: 10.1164/rccm.200502-235PP
Familial Lung CancerGenetic Susceptibility and Relationship to Chronic Obstructive Pulmonary DiseasePopulation Studies and Prevention Program, Karmanos Cancer Institute; and Department of Internal Medicine, Wayne State University, Detroit, Michigan Correspondence and requests for reprints should be addressed to Ann G. Schwartz, Ph.D., Karmanos Cancer Institute, 110 East Warren Avenue, Detroit, MI 48201. E-mail: schwarta{at}med.wayne.edu ABSTRACT Lung cancer continues to be the leading cause of cancer death, and although most lung cancer is attributable to cigarette smoking, underlying genetic susceptibility is suggested by studies demonstrating familial aggregation. The first family linkage study of lung cancer has identified linkage of lung, laryngeal, and pharyngeal cancer in families to a region on chromosome 6q2325. Because lung cancer and chronic obstructive pulmonary disease (COPD) are known to aggregate in families beyond shared risk associated with smoking, the linkage results are compared and contrasted with results from genomewide linkage and association studies and candidate gene studies searching for genes for lung cancer, lung function, and COPD. Linkage on chromosome 6q to both lung cancer and lung function, and on 12 to lung cancer, COPD, and lung function, together with overlap in candidate genes for these outcomes, suggests that future research into underlying genetic mechanisms of lung disease would benefit from broadening the collection of family history data and better defining the "high risk" population. As familial risk of lung disease is better defined, referral into screening programs and prevention trials can be better targeted to reach families with both a history of lung cancer and COPD.
Key Words: chronic obstructive pulmonary disease genetics linkage lung cancer Lung cancer remains at epidemic proportions and continues to be the leading cause of cancer death in both men and women (1). It is well accepted that 85 to 90% of all lung cancers are attributable to cigarette smoking (2), so that our major focus has been on prevention of initiation of smoking through youth-oriented programs and smoking cessation efforts. Through extraordinary personal and public health measures, the rate of smoking among adults has dropped substantially in the past two decades and there is now a concomitant drop in lung cancer rates (1). There remain, however, an estimated 46 million former smokers who continue to be at risk, in addition to the 45.8 million who continue to smoke (3). Because cigarette smoking is such an overwhelming risk factor, and preventable, the importance of family history and genetic susceptibility to lung cancer risk has often been overlooked. The first reports of familial aggregation of lung cancer were published over 40 years ago (4). Most observers have believed that the similar smoking environment within the family setting or workplace accounted for findings of familial aggregation, but even after adjusting for smoking patterns, individuals with a family history of lung cancer are at approximately two- to threefold increased risk of developing this disease (reviewed in References 57). This article attempts to place in context for the practicing pulmonary specialist and internist the role of familial risk in lung cancer and to suggest an important linkage to familial risk for chronic obstructive pulmonary disease (COPD) and the potential for shared genetic susceptibility. THE GENETIC EPIDEMIOLOGY OF LUNG CANCER Recently published work by Bailey-Wilson and coworkers (8) describes results of the first family linkage study searching for genes for lung cancer. This work was developed and conducted by the Genetic Epidemiology of Lung Cancer Consortium (GELCC), which includes a multidisciplinary team of investigators studying the genetics underlying lung cancer susceptibility. Linkage studies require the construction of large extended families, with multiple affected relatives from whom DNA can be obtained or genotypes inferred by genotyping spouses and offspring. Hundreds of genomewide markers are typed in family members to identify chromosomal regions that are inherited with disease. Although the evidence pointing to a gene for lung cancer is substantial, problems associated with the conduct of a linkage study in lung cancer are even greater. The average age of lung cancer diagnosis is 70 years and 5-year survival after a diagnosis continues to be poor, at 15%, so affected family members are typically deceased, as are their parents, siblings, and spouses. Lung tissue is often not available because not all lung cancer is detected at a resectable stage. The confounding risks due to variations in patterns of cigarette smoking also need to be considered. It took the multiinstitutional, multidisciplinary efforts of GELCC to conduct this study. Bailey-Wilson and colleagues (8) report that 13.7% of the over 26,000 patients screened with lung cancer had at least one first-degree relative with lung cancer. Fifty-two families were informative for linkage analysis, having at least three relatives with lung cancer, and were therefore included in the study. To allow for a more comprehensive look at the phenotype, family members with laryngeal and pharyngeal cancers were also considered to be affected. The mean age of onset for lung, laryngeal, and pharyngeal cancer in these families was 60 years. The median age of onset for the youngest family member affected in each family was 50.5 years, demonstrating a skewing toward an earlier age at onset within these high-risk families compared with the general population. More than 90% of those affected smoked cigarettes. Multipoint parametric linkage analysis was conducted using an autosomal-dominant, low-penetrance model assuming a susceptibility allele frequency of 0.01, 10% penetrance in gene carriers and 1% penetrance in noncarriers. The model used was based on segregation studies suggesting evidence for a dominant or codominant rare gene for lung cancer (reviewed in Reference 5). Under this simple model, a maximum HLOD (logarithm of the odds favoring genetic linkage given heterogeneity) of 2.79 on chromosome 6q (155 cM) was achieved. An LOD (logarithm of the odds favoring genetic linkage) score of 3.0 corresponds to a p value of 0.0001 and has traditionally been used as a significance threshold. In this study, HLOD scores were calculated (instead of LOD scores) because of the inclusion of African American and white families in whom underlying gene frequencies might vary; the HLOD analysis considers potential heterogeneity. Recent work suggests that using an HLOD of 4.0 as the critical threshold for declaring significance is conservative when multi- point HLODs are calculated and multiple penetrance models are used (9). In a subset of 38 families with four or more affected family members, the HLOD was 3.47. In the most highly affected families, with five or more affected relatives in multiple generations (n = 23), the HLOD was 4.26, with 94% of these families estimated to be linked to this region. The 1-HLOD support interval in these 23 families extended from 146 to 164 cM. The reported HLOD in these most highly affected families exceeds the conservative critical significance threshold. Nonparametric tests, which do not require specification of an underlying inheritance model, supported the parametric linkage findings. This work demonstrates the existence of a chromosomal region linked to lung cancer inheritance in families. This work also provides some initial insight into the relationship between smoking and lung cancer occurrence in high-risk families. In a set of 21 of the 23 families most tightly linked to this region, carrier status was assigned. In both carriers and noncarriers of a putative lung cancer susceptibility region, smoking was associated with a threefold increase in lung cancer risk. However, it was only among the noncarriers that the typical doseresponse relationship was observed, with a 2% increase in lung cancer risk with each pack-year of exposure (p = 0.0023). Among the carriers, this doseresponse relationship was not apparent, indicating that any amount of smoking increases risk in this group. These results provide the first evidence supporting linkage between a region on 6q2325 (146164 cM) and lung, laryngeal, and pharyngeal cancer. Additional analyses are being conducted that are limited to lung cancer. Allelic loss in this region has been reported in a number of tumor types, suggesting the existence of a tumor suppressor gene. The region identified is large, encompassing 74 known genes and 41 unknown genes, including four putative tumor suppressor genes (SASH1, LATS1, IGF2R, and PARK2), as well as genes involved in regulating cellular proliferation and preventing DNA damage. In addition to linkage on 6q, some evidence for linkage on 12q was also reported. Expanding the number of families studied and restricting analyses to lung canceraffected members only are necessary next steps. The work to identify the specific gene(s) associated with familial lung cancer is just beginning. An alternative approach to lung cancer gene identification was taken by Yanagitani and colleagues (10, 11). Adenocarcinoma of the lung in mice has been mapped to mouse chromosome 6q near the Kras2 locus. This region also contains the pulmonary adenoma susceptibility 1 (Pas1) locus linked to inherited predisposition to lung adenocarcinoma in mice. Kras2 and Krag (K-ras oncogene-associated gene) loci are within the Pas1 region. In the human, Kras2 and Krag have been mapped to 12p12.1 and 12p11.2, respectively. Using 13 microsatellite markers in the 12p1212q12 region and comparing cases with adenocarcinoma of the lung with control samples, Yanagitani and coworkers (10) reported increased risk of lung cancer associated with marker D12S0134, suggesting a putative lung adenocarcinoma locus in this region. This marker falls between Kras2 and Krag. Yanagitani and colleagues (11) also performed a genomewide association study using 322 microsatellite markers in a series of cases with adenocarcinoma of the lung and hospital-based control subjects. Cases and control subjects differed in allele and genotype frequencies at D6S474 at 6q22 (p = 0.06), close to the family linkage region. A significant association was also seen with marker D19S246, at 19q13.3 (p = 0.03). In particular, the distribution of A1, A8, and A9 alleles was significantly different between cases and control subjects after adjusting for age, sex, and smoking. ERCC1 and XPD, DNA repair genes, are located in this region at 19q13.23. Because control subjects in this study included individuals with chronic lung diseases, risk estimates from this study are likely to be underestimated if similar pathways underlie risk of lung cancer and chronic lung diseases. Although a specific lung cancer gene has not been identified, understanding risks associated with family history of lung cancer is essential in the clinical setting. One recent study estimated lung cancer risk given family history and smoking history (7). Estimates such as these can be used in counseling patients and targeting groups for inclusion in screening and prevention trials. In addition to smoking and family history of cancer, the link between an obstructive lung disease diagnosis and risk of lung cancer needs to be reevaluated. Airway obstruction has been shown prospectively to increase risk of developing lung cancer (12), as has COPD (13), and has the same underlying risk profile: about 10 to 20% of smokers develop COPD and there is familial aggregation of COPD and lung function. THE LUNG FUNCTION, COPD, AND LUNG CANCER CONNECTION
Given the similarities and association between COPD and lung cancer, an obvious question is whether there is a common underlying genetic susceptibility acting in addition to the known shared risk associated with cigarette smoking. As far back as the mid-1970s, Cohen and colleagues (14, 15) demonstrated a common familial component to lung function, COPD, and lung cancer not completely explained by smoking or LINKAGE STUDIES Although the familial lung cancer linkage study was the first of its kind, a number of linkage studies have been done to identify genes for lung function and COPD. In a linkage study undertaken using 330 families in the Framingham study, Joost and coworkers (17) identified chromosomal regions influencing FEV1 and FVC. This study, like the GELCC linkage study, used 399 markers spaced at an average 10 cM across the genome. A region on 6q was most strongly linked to FEV1, with an LOD score of 2.4. The region of tightest linkage was at the 6q terminus, with increasing LOD scores beginning at approximately 165 cM, just beyond the newly described lung cancer linkage region, and extending to the end of the chromosome. In an effort to narrow the region, additional markers were typed, resulting in an LOD score of 5.0 at 184.5 cM (18). This study was conducted in a population unselected for lung disease and the authors suggest that a gene in this region is most likely to influence FEV1 and airflow obstruction. A candidate gene in this region is SMOC2 (secreted modular calcium-binding protein 2). The encoded protein contains a serine protease inhibitor domain.
A similar linkage study in 72 families ascertained through severe early-onset COPD probands without severe The linkage studies conducted to date have been relatively small, regions identified are large and include a number of genes, and regions do not exactly overlap between studies. Kruglyak and Lander (22), however, suggest that, because of underlying linkage disequilibrium, susceptibility genes will often lie outside regions of maximum allele sharing, particularly when genes confer risk on the order of two- to threefold. The criteria used to define statistical significance of linkage studies (23) are also debatable, as is the best approach (i.e., parametric vs. nonparametric). Even with these limitations, there is some potential overlap in regions linked to lung function, COPD, and lung cancer on chromosomes 6q and 12p. These linkage studies can serve as initial screens for more focused mapping and experimental approaches to prove causality. CANDIDATE SUSCEPTIBILITY GENES FOR COPD AND LUNG CANCER Alternative approaches to gene identification include candidate gene association studies, and work in this area also suggests possible overlap between genes predicting lung function, COPD, and lung cancer. Candidate genes in several groups are discussed below and presented in Table 1.
1-Antitrypsin is an antiprotease that binds and inhibits neutrophil serine proteases such as elastase in the lung, protecting against lung tissue destruction. An imbalance between neutophil elastase and 1-antitrypsin (or, more generally, a proteinase/antiproteinase imbalance) has been hypothesized to contribute to the development of COPD, as well as lung cancer (24). The gene for 1-antitrypsin is located on chromosome 14q32.1 and over 75 alleles have been identified. Two common alleles (S and Z) are associated with 1-antitrypsin deficiency. It is well known that individuals homozygous for the Z allele are at high risk of COPD. The deficiency results in unneutralized neutrophil elastase and an associated breakdown of elastin in lung tissue, leading to early-onset and severe emphysema. Cigarette smoking contributes to faster tissue destruction through stimulation of neutrophils and increased secretion of elastase and further inactivation of 1-antitrypsin. Recent studies have shown that the MZ genotype is also associated with decline in lung function (25, 26) and airway obstruction (27), whereas one study reported no association (Table 1) (28). A recent meta-analysis concludes that the MZCOPD association is stronger in case-control studies (odds ratio (OR), 2.97; 95% confidence interval (CI), 2.084.26) than in cross-sectional studies (OR, 1.50; 95% CI, 0.972.31) (29). Both Z and S allele carriers have been reported to be more common among patients with lung cancer than in the general population (30), but only one study has been completed. Neutrophil elastase itself has been evaluated as a candidate in lung cancer etiology. The gene encoding this enzyme (ELA2) is located on chromosome 19p13.3. Taniguchi and colleagues (31) identified two polymorphisms in the promoter (T-903G [REP-a] and G-741A [REP-b]) and evaluated the contribution to lung cancer risk from these polymorphisms in a case-control setting. The TT genotype at REP-a was associated with 2.3-fold increased risk of lung cancer (95% CI, 1.24.7), whereas the GG genotype at REP-b was associated with a 1.4-fold increased risk (95% CI, 1.02.0). Higher promoter activity was associated with the risk genotypes. Similar findings for the TT genotype at REP-a were reported by Park and coworkers (OR, 3.2; 95% CI, 1.0310.4) (32). The role of neutrophil elastase in COPD development has not been evaluated. In addition to breaking down lung tissue, neutrophil elactase activates matrix metalloproteinases (MMPs), a family of more than 20 proteolytic enzymes that degrade extracellular membranes. MMP1 (interstitial collagenase), MMP12 (human macrophage elastase), and MMP9 (gelatinase B) have been associated with lung disease in animal models. Haplotypes at MMP1/MMP12 and an MMP1 promoter single nucleotide polymorphism (SNP), G-1607GG, have been associated with lung function decline in smokers (Table 1) (33). The same MMP1 promoter SNP alters risk of lung cancer (34), as does an MMP3 promoter SNP (35). In addition, lung cancer risk associated with the MMP1 2G/2G genotype has been shown to be higher in those with emphysema than in those without this condition (36). An MMP2 promoter SNP is associated with a twofold increase (95% CI, 1.72.8) in lung cancer risk (Table 1) (37). The MMP genotypelung cancer risk association was stronger as cigarette smoke exposure increased (34, 37).
COPD is characterized by inflammation of the airway induced by cigarette smoke. Animal models implicate IL-13 overexpression in MMP- and cathepsin-dependent emphysema, elevated mucus production, and COPD-like inflammation. SNPs occur in IL-13 (5q31) and the IL-13 receptor, made up of an IL-4 receptor Microsomal epoxide hydrolase is found in bronchial epithelial cells and hydrolyses arenes, alkenes, and aliphatic epoxides, making them less reactive. However, in some instances, polycyclic aromatic hydrocarbons found in tobacco smoke become more reactive when metabolized by microsomal epoxide hydrolase. The gene for microsomal epoxide hydrolase, EPHX1, has been mapped to 1q42.1 and includes two polymorphisms: Try113His, which is associated with decreased activity, and His139Arg, which is associated with increased activity. The low-activity haplotype (His113His139) has been associated with rapid decline in lung function (26), whereas the His113 genotype has been associated with COPD (4447). Studies in lung cancer have generally reported that the predicted high-activity genotype is associated with increased risk (Table 1) (4855). In one of the largest studies, lung cancer risk associated with EPHX1 varied with pack-years of smoking, with a twofold increased risk among nonsmokers with the low-activity genotype and 30% decreased risk in smokers of 80 pack-years with the low-activity genotype (56). This decease in lung cancer risk in individuals with the low-activity genotype (or equivalently, increased risk in those with the high-activity alleles) was similar to that seen in a pooled analysis (51). Given microsomal epoxide hydrolase's role in detoxification and activation, it is not surprising to see variation in results based on cigarette smoke exposure and by disease outcome. CONCLUSIONS The prevention of lung cancer through the elimination of smoking continues to be the best way to decrease mortality from this disease in the long term. However, even if everyone stopped smoking today, lung cancer incidence and mortality would continue to be high for several decades. While debate continues about the efficacy of spiral computed tomography screening for lung cancer in broad populations of smokers, the ability to focus screening efforts in a truly high-risk subpopulation would clearly be of benefit now. Pack-years of exposure and lung function have been used as criteria for entry into screening trials and chemoprevention studies. Consistent epidemiologic data and the recent lung cancer linkage results suggest that family history should be included in a "high risk" definition as well. We are of the opinion that studies of spiral computed tomography screening and lung function testing is warranted in current or former smokers with a strong family history of lung cancer, particularly when age at diagnosis of the family member is young (7). Associations between markers on chromosome 6q to both lung cancer and lung function, and on 12 to lung cancer, COPD, and lung function, together with overlap in candidate genes for these outcomes, suggest that future research into underlying genetic mechanisms of lung disease would benefit from broadening the collection of family history and lung function data. We suggest that the collection of family history of lung cancer in patients with COPD and family history of COPD in patients with lung cancer become routine. Age at diagnosis of family members with lung cancer or COPD should be collected, if possible. As familial risk of lung disease is better defined, referral into screening programs and prevention trials can be better targeted to reach families with both a history of lung cancer and COPD. The identification of genes for lung cancer and COPD will require larger linkage and association studies that would benefit from the collection of lung function and lung disease history and a more extensive family history. FOOTNOTES Supported by National Institutes of Health grants R01 CA60691, U01 CA76293, and P30 CA22453-24, and contract N01-PC-25005. Originally Published in Press as DOI: 10.1164/rccm.200502-235PP on September 1, 2005 Conflict of Interest Statement: Neither of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form February 14, 2005; accepted in final form August 31, 2005 REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||