The methylation status of Wnt XL184 antagonist genes including SFRP1, SFRP2, SFRP5, WIF1, DKK3, APC, and CDH1, defined as their epigenotype, was detected by Methylation Specific PCR Assays (examples were shown in Additional file 1: Figure S1A). The frequency of methylation events in Wnt antagonist genes in patients with different demographic characteristics was listed in Table 1. Interestingly, no significant difference in epigenotype of Wnt antagonist

genes was found between male and female, among different age groups, between smokers and non-smokers, or between adenocarcinoma and non-adenocarcinoma cases. Using DHPLC, we also detected EGFR activating mutations in exon 19 or 21 (the examples of wild type, mutated exon 19, and mutated exon JQEZ5 mw 21 were shown in Additional file 1: Figure S1B, 1C, and 1D). Among the 155 patients, 85 (55.4%) carried mutations in either exon 19 or 21 of the EGFR genes (Table 1).Similar to the previous studies, we found that EGFR mutation rates were significantly increased

among the patients younger than 65 years old (P = 0.02, Fisher’s exact test) and the patients who are nonsmokers (P = 0.04, Fisher’s exact test). EGFR mutation reversely correlates with sFPR1 methylation (P = 0.005) and sFRP5 (P = 0.011). We fail to find methylation of other wnt antagonist genes correlated with EGFR mutation (Table 2). Table 2 P value among methylated genes and EGFR mutation   sFRP1 sFRP2 sFRP5 DKK3 WIF-1 APC CDH-1 EGFR mutation sFRP1 NA 0.004 MAPK inhibitor 0.005 0.008 0.02 <0.0001 0.266 0.005 sFRP2 0.004 Janus kinase (JAK) NA <0.0001 <0.0001 0.007 <0.0001 <0.0001 0.854 sFRP5 0.005 <0.0001 NA <0.0001 <0.0001 0.06 <0.0001 0.011 DKK3 0.008 <0.0001 <0.0001 NA 0.0001 0.006 <0.0001 0.489 WIF-1 0.02 0.007 <0.0001 <0.0001 NA 0.03 0.02 0.094 APC <0.0001 <0.0001 0.06 0.006 0.03 NA 0.126 0.546 CDH-1 0.266 <0.0001 <0.0001 <0.0001 0.02 0.126 NA 0.592 EGFR 0.005 0.854 0.011 0.489 0.094 0.546 0.592 NA mutation                 We next investigated whether

the epigenotype of any Wnt antagonist genes correlated with the genotype of EGFR. Hierarchical clustering of the epigenotype of SFRP1, SFRP2, SFRP5, WIF1, DKK3, APC, and CDH1, as well as the genotype of EGFR (defined as “1” if mutation was detected in the exon 19 or 21, and as “0” if no mutation was detected) was generated using Partek Genomics Suite 6.5 (Partek Inc., MO). As shown in Figure  1, the epigenotype of Wnt antagonist genes had similar patterns, which were different from the genotype of EGFR. Therefore, our results suggested that the DNA methylation of Wnt antagonist might be independently regulated from the genotype of EGFR. Figure 1 Hierarchical clustering of Wnt antagonist DNA methylation status and EGFR genotype in 155 patients received EGFR-TKI therapy.