The Role of Phase II Detoxification Genes Towards Modulating Risk for Lung Cancer and its Association with Clinical Outcomes

Abstract

Background: Lung cancer is the leading cause of cancer mortality globally, and the key risk factors are smoking and occupational exposure. The presence of polymorphic genes affecting the levels of metabolic activation and carcinogen detoxification influences lung cancer susceptibility. Objectives To evaluate the role of single nucleotide polymorphic variants of Phase II detoxification genes NQO1 (rs1800566), SULT1A1 (rs9282861), EPHX1 (rs1051740, rs2234922), NAT2 (rs1799929, rs1799930, rs1799931, rs1208), MTHFR (rs1801133, rs1801131) and GSTP1 (rs1695). Methodology Genotyping of genomic DNA was carried out using PCR-RFLP (polymerase chain reaction-restriction fragment length polymorphism) for each polymorphic site under study. Total number of subjects genotyped in case of NQO1, SULT1A1, EPHX1, and NAT2 were 1100 (550 cases, 550 controls); in MTHFR were 253 cases and in case of GSTP1 682 cases, respectively. Following this association study, logistic regression was used to get the adjusted odds ratio and significance. To find the possible connection between interacting SNP-SNP and gene-gene interactions, data mining analysis was performed that included both Multi-dimensionality reduction (MDR) and Classification and Regression tree (CART) analysis. Overall survival analysis was carried out using Kaplan-Meier survival analysis and Cox-regression analysis, which yielded the adjusted hazards ratio. Results The TT genotype for NQO1 609C>T polymorphism exhibited a good association with LC risk in regards of clinical-pathological parameters. A 3.5 fold higher odds in subjects with stageIII (p=0.0006),5 fold higher probability of lymph-node invasion(p=0.007) and an odds of less than one in case of metastasis(p=0.0028). Furthermore, 473 subjects were analyzed in regards of overall survival, wherein the TT genotype exhibited a better OS than CC genotype (9.43vs.8.13months). Whereas, none of the chemotherapeutic regimen exhibited a significant association with OS, except for the subjects possessing TT genotype and administered with paclitaxel+cis/carboplatin, these exhibited a very poor survival(3.57vs.12.20months).In such patients a significant higher death rate was obtained using Cox-proportional analysis(HR’=7.95;p=0.0098). For SULT1A1 Arg213His polymorphism the variant genotype exhibited no association with LC risk, even after stratification on basis of histological subtypes. The male LC patients carrying the variant His213 allele (p=0.02) did not exhibited an increased risk towards LC. The smokers harboring either the Arg/His (p=0.019) or Arg/His+His/His (p=0.01) genotype, exhibit a reduced risk towards LC. Furthermore, the LC subjects who were heavy smokers and harboring the Arg/His (p=0.01) genotype did not show a genetic predisposition towards LC susceptibility. The subjects who smoked pack years of above 40 and carrying the His/His (p=0.036) and Arg/His+His/His (p=0.015) genotype were found to pose no effect on the risk of LC. Furthermore, 473 subjects were analyzed in regards of overall survival, wherein the His/His genotype exhibited better OS than Arg/Arg genotype (11.30 vs.8.07months). Variant genotype (His/His) of Tyr113His polymorphism exhibited a two-fold increased odds of lymph node invasion (p=0.04). In EPHX1 Tyr113His polymorphism, the Tyr/His (p=0.01) and Tyr/His+His/His (p=0.01) genotype was a risk factor for smokers. Further, subjects harboring His/His (p=0.04) and Tyr/His (p=0.04) genotype in case of light and heavy smokers, respectively were found to be associated with an increased risk of LC. Further, the Tyr/His (p=0.03) and Tyr/His+His/His (p=0.03) genotype were found to be associated with risk of LC in male subjects. The subjects carrying the combined genotype for His139Arg (His/Arg+Arg/Arg) exhibited a better OS using Cox-proportional hazards (9.80 vs.7.60months, p=0.04) and heterozygous genotype exhibited a better OS in case of SCLC(11.30 vs.6.73months,log rank test p=0.02). Further, it was found that heterozygous genotype of EPHX1 Tyr113His gene had a longer survival time in those patients who were receiving cisplatin/carboplatin along with irinotecan(11.30 vs.7.23months,log rank test p=0.007). The NAT2 857G>A polymorphism exhibited a risk towards lung cancer (p=0.005). Whereas, variant alleles for the 481C>T polymorphism had a decreased risk for LC (p=0.0003). Further, NAT2*4/*7 polymorphism conferred a positive association between genotype and ADCC (p=0.001) and 481C>T polymorphism had a decreased risk for SQCC (AOR=0.39, p=0.0006) and SCLC (p=0.001) subjects. The smokers carrying mutant genotype for the 481C>T polymorphism had a decreased risk towards LC (p<0.0001) even in light (p=0.002) as well as heavy smokers (p=0.001). In case of female, 2.59 folds and 3.66 folds increased risk of LC development was observed in subjects with intermediate and slow acetylator for the 857G>A polymorphism. Whereas, in case of males this polymorphism depicts a reduced risk for lung cancer. On the other hand, 803A>G depicted a 2.82 folds risk of lung cancer in case of female subjects who were slow acetylators. Our study exhibits a significant difference in the overall haplotype distribution between cases and controls. In our study overall, (857G>A, 481C>T, 803A>G) was found to be best model, but was not significant using MDR. Considering the CART results 481C>T polymorphism came out to be the most significant factor in determining the LC risk. For the 803A>G polymorphism a 3-fold odds of lymph node invasion were observed for mutant genotype, the recessive model exhibited an odd of 2.8. 590G>A appears to be a potential prognostic factor for OS of SCLC patients after irinotecan therapy as the survival time for such patients was better. The patients who were carrying both the mutant alleles (C/C) for the MTHFR (1298A>C) polymorphism showed a higher trend of median survival time (MST) as compared to the patients bearing the wild type genotype (A/A) (MST=13.93 vs. 7.97, p=0.12). On basis of toxicity profiling, we observed that lung cancer patients with the AT (heterozygous) genotype for MTHFR A1298C polymorphism had a 12-fold risk of diarrhoea (AOR=12.54, 95% CI=1.54-101.86, p=0.018). The patients with the heterozygous genotype (CT) of the MTHFR C677T polymorphism had a 5.34-fold increased risk of developing neutropenia toxicity (AOR=5.34, 95% CI= 1.49-19.06, p=0.009). On basis of toxicity profiling, we observed that lung cancer patients with mutant genotype of GSTP1 Ile105Val had an increased risk of leukopenia (OR=2.41; 95%CI=1.39-4.18, p=0.001). Similarly, data also suggested that that patients with the heterozygous genotype (Ile/Val) had a 2.14-fold increased risk of developing severe anemia (OR=2.14, 95% C.I.=0.97-4.62, p=0.03). Further our data also showed that in SCLC patients polymorphism of GSTP1 was associated with thrombocytopenia (χ2 test=7.32, p=0.02). The five-factor model (NQO1 P187S, EPHX1 Y113H, NAT2*6 R197Q, NAT2*5B L161L, NAT2*5C K268R) was the overall best model for the interaction of selected Phase II polymorphisms. The CART results demonstrates that NAT2 481C>T caused the initial split and genotypic combination of NAT803A>G(M)/NAT857G>A(M)/EPHX1337T>C(W)/NQO1609C>T(M)/EPHX1415A>G(W)/ NAT590 G>A(W)/SULT1A1638G>A(W)/NAT481C>T(M) conferred the highest risk towards lung cancer (OR=24.35, 95% C.I.=2.94-201.26, p= 0.003). In terms of analyzing toxicity profile, the mutant genotype for NQO1 exhibits reduced risk for gastrointestinal toxicity, including constipation and anorexia. Similarly, the heterozygous and combined genotype for SULT1A1 polymorphism had decreased risk for developing neutropenia. Further, the heterozygous and combined genotype for NAT2*6 polymorphism was found to be associated with an increased risk for leukopenia and anemia, i.e., hematological toxicity as well as neutropenia. At the same time, NAT2*5C was associated with a reduced risk of developing hematological toxicity, especially anemia. The heterozygous genotype for NAT2*7 was found to be associated with 2 folds increased risk of thrombocytopenia but decreased risk of anemia (hematological toxicities) as well as decreased risk of alopecia and anorexia. Conclusion These results suggest that NQO1 variant genotype may modulate LC risk, especially those with stage III LC and as poor prognostic factor. The SULT1A1 Arg213His polymorphism provides no epidemiologic evidence of genetic predisposition towards LC risk in relation to tobacco smoking. The EPHX1 Tyr113His polymorphism is associated with LC risk in smokers and appears to be a potential prognostic factor for OS of SCLC patients after irinotecan therapy and might increase the likelihood of lymph node metastasis. Whereas, the EPHX1 His139Arg exhibit better survival especially, in SCLC patients. The NAT2 variant genotype for 590G>A and 803A>G was not found to modulate risk towards LC, but 857G>A polymorphism exhibited a risk towards LC and 481C>T polymorphism had a decreased risk for LC. NAT2 590G>A appears to be a potential prognostic factor for OS of SCLC patients after irinotecan therapy and 480C>T came out to be significant factor using CART. Our results suggest that MTHFR polymorphism is associated with severe hematological toxicity like leukopenia and also with gastrointestinal toxicity. Thus, evidence suggests that MTHFR polymorphism might have an impact on the development of pemetrexed and platinum related toxicities but not as a clinical predictor of efficiency or as a prognostic marker. Our results suggest that GSTP1 Ile105Val polymorphism could act as a predictive biomarker for certain hematological toxicity, like leukopenia and anaemia, but not thrombocytopenia or neutropenia.