**3.2.1** *XPD* **Lys751Gln and** *XPD* **Asp312Asn**

The *XPD* Lys751Gln polymorphism is the most frequently studied of the NER polymorphisms in association with risk of cancer. In our Danish prospective study on the Diet, Cancer and Health cohort, we observed no association of the *XPD* Lys751Gln and *XPD* Asp312Asn polymorphisms with risk of colorectal cancer (Hansen et al., 2007). Previously, several studies had similar findings of no association between the *XPD*  Lys751Gln (Moreno et al., 2006; Huang et al., 2006; Berndt et al., 2006; Mort et al., 2003; Starinsky et al., 2005; Skjelbred et al., 2006b; Engin et al., 2010; Stern et al., 2009; Stern et al., 2007; Yeh et al., 2005; Joshi et al., 2009; Wang et al., 2010) and the *XPD* Asp312Asn (Moreno et al., 2006; Huang et al., 2006; Berndt et al., 2006; Goodman et al., 2006; Stern et al., 2009; Stern et al., 2007; Joshi et al., 2009) polymorphisms and risk of colorectal cancer. Additionally, Bigler and colleagues found no association of the two polymorphisms with development of adenomas (Bigler et al., 2005). However, they detected a higher risk of colorectal adenomas among individuals with at least two variant alleles of the *XPD* polymorphisms, with an OR of 1.57 (CI: 1.04-2.38). When stratifying by age the association of the two polymorphisms with risk of adenomatous polyps was restricted to the individuals younger than 60 years when diagnosed (OR=3.77, CI: 1.94-7.35). The risk of adenomatous polyps was higher among smokers carrying the homozygous *XPD* variant alleles (OR=3.93, OR: 1.68-9.21) compared with non-smokers carrying the homozygous wild type. A similar finding could not be detected on risk of hyper-plastic polyps. In our Danish study (Hansen et al., 2007) and in a Singapore Chinese study (Stern et al., 2007) did neither of the two XPD polymorphisms, *XPD* Lys751Gln or *XPD* Asp312Asn, modify the effect of smoking on risk of colorectal cancer.

Goodman *et al.*, did not detect any SNP-SNP interaction between the *XPD* Asp312Asn polymorphism and other NER polymorphisms (Goodman et al., 2006). Skjelbred and colleagues detected an association between the *XPD* Lys751Gln polymorphism and development of colorectal adenomas, with an OR of 1.40 (CI: 1.08-1.81), among carriers of the variant allele compared to carriers of the homozygous wild type allele (Skjelbred et al., 2006b). The statistical significance was limited to the low-risk adenoma group (OR: 1.46, CI: 1.11-1.90). The results were contradicted by a large study by Stern *et al.,* including 740 cases with adenomas and 789 controls, where a lower risk of adenomas was observed (OR=0.7, CI:

Polymorphisms in Nucleotide Excision Repair

associated with colorectal cancer risk.

*XPA* (Berndt et al., 2006) with risk of colorectal cancer.

suggest no association with risk of colorectal cancer.

with risk of colorectal cancer (Mort et al., 2003; Berndt et al., 2006).

**3.3 SNPs in NER and risk of other types of cancer than colorectal cancer** 

et al., 2009) .

**3.2.3** *XPA* **G23A** 

**3.2.4** *ERCC1* **Asn118Asn** 

et al., 2004; Hu et al., 2006).

Genes and Risk of Colorectal Cancer: A Systematic Review 547

In a large American study by Huang three polymorphisms in *XPC* was studied, including the *XPC* Lys939Gln polymorphism. No association was found between the *XPC* Lys939Gln polymorphism and risk of adenomas (Huang et al., 2006). However, higher risk for development of adenomas was observed among current or recent smokers carrying the *XPC* 939Gln allele (OR=2.0, CI: 1.3-3.0) or a *XPC* haplotype encompassing three linked SNPs in *XPC* (Arg492His, Ala499Val, Lys939Gln) compared with neversmokers carrying the homozygous wild type allele. A study by Joshi *et al*. observed no association between the *XPC* intron 11 polymorphism and risk of colorectal cancer (Joshi

In a small study by Berndt *et al*. a tendency for higher risk of proximal colon cancer was observed among homozygous carriers of the variant *XPC* Lys939Gln allele, with an OR of 1.74 (CI: 0.98-3.08) (Berndt et al., 2006). The result may possibly be a chance finding due to sample size and multiple testing. Three other SNPs in the *XPC* gene, see Table 2, were not

To our knowledge, only three studies have been published on the association of polymorphisms in the *XPA* gene with risk of colorectal cancer: The studies by Berndt *et al*., Joshi *et al*., and our study. For a polymorphism positioned in the *XPA* 5´ UTR region, a lower risk for colon cancer cancer was observed among carriers of the T-allele (OR=0.4, 95% CI: 0.2-0.8) compared with homozygous carriers of the C-allele (Joshi et al., 2009) . There was no association for risk of rectal cancer. No association was observed of the *XPA* G23A polymorphism (Hansen et al., 2007) or a polymorphism in the 3´ un-translated region of

The results from studies by Skjelbred *et al*. (Skjelbred et al., 2006a) , Joshi *et al*. (Joshi et al., 2009) , and our Danish study (Hansen et al., 2008) on the *ERCC1* Asn118Asn polymorphism

Moreno *et al*. examined five polymorphisms in the *ERCC1* gene. A haplotype containing the minor allele of three of the *ERCC1* polymorphisms was associated with a higher risk of colorectal cancer (OR=2.3, 95% CI: 1.0-5.3) compared with carriers of the most frequent haplotype (Moreno et al., 2006). Two other SNPs in the *ERCC1* gene were not associated

Numerous association studies of polymorphisms in genes involved in NER are reported on various types of cancer, with the majority of studies focused on the *XPD* Lys751Gln and *XPD* Asp312Asn polymorphisms. A meta-analysis of lung cancer by Kiyohara *et al.* (with 1913 cases and 1882 controls of different ethnicities) (Kiyohara & Yoshimasu, 2007) suggested among other studies (Xing et al., 2002; Hu et al., 2004; Yin et al., 2006), that carriers of the variant alleles of either of the two *XPD* polymorphisms were found to be at higher risk of lung cancer, while a number of other studies did not observe any association of the two polymorphisms with lung cancer risk (De et al., 2007; Vogel et al., 2005b; Popanda

0.4-1.0) among homozygous carriers of the *XPD* 751Gln allele (Stern et al., 2006). The result was not stratified for ethnicity (Caucasian, African-American, Latinos, Asian-Pacific Islander). When excluding the 1 case and the 17 controls of Latinos, the OR increased to 0.9 (confidence intervals were not reported). An interaction between the *XPD* Lys751Gln polymorphism and alcohol consumption was observed (P=0.04), with higher risk of adenomas among ever-drinkers carrying the *XPD* 751 Gln/Gln genotype (OR=2.5, CI: 1.2- 5.2) compared with never-drinkers carrying the same genotype. There was no interaction between the polymorphisms *XPD* Lys751Gln or *XPD* Asp312Asn, respectively, and alcohol consumption on risk of colorectal cancer in our Danish study (Hansen et al., 2007) and in the Singapore Chinese study (Stern et al., 2007).

In a family-based case-control study using a case-only design, an interaction was observed between the two polymorphisms, *XPD* Lys751Gln and *XPD* Asp312Asn, and intake of heavily browned red meat on colorectal cancer risk (Joshi et al., 2009). Intake of red meat heavily brown on the outside or inside increased the risk for colorectal cancer only among carriers of the *XPD* codon 751 Lys/Lys genotype or the *XPD* codon 312 Asp/Asp genotype (case-only interaction P <0.006). There was no association between the meat intake and colorectal cancer risk when the individuals carried at least one copy of the Asn321 or Gln751 alleles. The results remained statistically significant after accounting for multiple testing. No interaction was observed in our Danish study between the two *XPD* polymorphisms and intake of red meat on risk of colorectal cancer (Hansen et al., 2007).

A higher risk of colorectal cancer has been observed among Ashkenazi Jews below 50 years of age when diagnosed (Starinsky et al., 2005). The risk was higher among carriers of the *XPD* 751Gln allele, but it may be a chance finding due to low number of cases (only 15 cases were diagnosed before their 50 years birthday). Furthermore, the Ashkenazi population is known to have particular genetic characteristics, why the result may not be generalized to other populations.

A large study from Taiwan observed a non-significant tendency for higher risk of colorectal cancer among men carrying the *XPD* 751Gln allele (OR=1.5, CI: 0.9-2.3), while no association was observed for women (OR=0.9, CI: 0.6-1.5) (Yeh et al., 2007). A similar tendency for a gender specific effect of the *XPD* Lys751Gln polymorphism was observed in our Danish study, with lower risk of colorectal cancer among women carrying the variant allele of *XPD* Lys751Gln with an IRR less than 0.62 among carriers of the *XPD* 751Gln allele, compared to women carrying the wild type allele (Hansen et al., 2007) . No association was found among men. The gender differences could hypothetically be caused by a hormonal interaction. However, we observed no interaction between the use of hormone replacement therapy among women and the polymorphism. Thus, we did not find the hypothesis plausible and conclude that our result in the Danish study may be a chance finding.

#### **3.2.2** *XPC* **Lys939Gln**

In our Danish study and in a Turkish study by Engin *et al*. (Engin et al., 2010) , the *XPC* Lys939Gln polymorphism was not associated with risk of colorectal cancer (Hansen et al., 2007). However, we did observe an interaction between the polymorphism and intake of red meat, with an IRR of 3.70 (CI: 1.70-8.04) for colorectal cancer per 100g red meat intake per day among homozygous carriers of the *XPC* Lys939Gln variant allele (Hansen et al., 2007) . In the light of the sample size and the multiple comparisons being made, this result may be a chance finding. The association was not statistically significant after a Bonferroni correction.

In a large American study by Huang three polymorphisms in *XPC* was studied, including the *XPC* Lys939Gln polymorphism. No association was found between the *XPC* Lys939Gln polymorphism and risk of adenomas (Huang et al., 2006). However, higher risk for development of adenomas was observed among current or recent smokers carrying the *XPC* 939Gln allele (OR=2.0, CI: 1.3-3.0) or a *XPC* haplotype encompassing three linked SNPs in *XPC* (Arg492His, Ala499Val, Lys939Gln) compared with neversmokers carrying the homozygous wild type allele. A study by Joshi *et al*. observed no association between the *XPC* intron 11 polymorphism and risk of colorectal cancer (Joshi et al., 2009) .

In a small study by Berndt *et al*. a tendency for higher risk of proximal colon cancer was observed among homozygous carriers of the variant *XPC* Lys939Gln allele, with an OR of 1.74 (CI: 0.98-3.08) (Berndt et al., 2006). The result may possibly be a chance finding due to sample size and multiple testing. Three other SNPs in the *XPC* gene, see Table 2, were not associated with colorectal cancer risk.
