Patients included in the study were pathologically confirmed to have lung cancer at Konkuk University Hospital (Seoul, South Korea) between July 2006 and December 2008. Patients were prospectively observed for tumor response and survival outcome. The present study was reviewed and approved by the Institutional Ethic Committee at Konkuk University Hospital. All patients provided informed consent to participate in the study and permission for use of tumor tissues.

Areas of the tissue and cytology slides containing the tumor cells of interest were marked by a cytopathologist using a pen. A diamond-tipped pencil was then used to mark the underside of the slide. Following removal of the cover slip, tumor cells were scraped with a 26-gauge needle. Briefly, 50–100 μl of DNA extraction buffer solution (50 mM Tris buffer, pH 8.3; 1 mM EDTA, pH 8.0; 5% Tween-20 and 200 μg/ml proteinase K) with 10% resin was added to the scraped cells. Following incubation at 56°C for at least 1 h, the tube was heated to 100°C for 20 min followed by centrifugation at 12,000 rpm for 10 min at 4°C to pellet the debris. The recovered supernatant was used for the PCR.

Biotinylated (B) PCR primer sequences for the amplification of EGFR mutation sites were as follows: exon 18, 5′-B-GCTCCC AACCAAGCTCTCTT-3′(F) and 5′-TATACACCGTGCCGA ACGC-3′(R); exon 19, 5′-GCATGTGGCACCATCTCA-3′(F) and 5′-B-AAAAGGTGGGCCTGAGGTT-3′(R); exon 20, 5′-B-ATGGCCAGCGTGGACAAC-3′(F) and 5′-TTTGTG TTCCCGGACATAGTC-3′(R); exon 21, 5′- ACCGCAGCA TGTCAAGATCAC-3′(F) and 5′-B-TCCGCACCCAGC AGTTTG-3′(R). In this method, 5 μl of DNA was added to produce a 50 μl of PCR solution mixture that contained 0.2 mmol of each dNTP, 1.5 mmol/l MgCl₂, 1X PCR buffer, 1.5 units of immolase DNA Taq polymerase and 20 pmol of each primer. PCR was performed with an initial denaturation for 5 min at 95°C followed by 40 cycles of 30 sec at 95°C, 30 sec at 54°C (exons 18, 21), 60°C (exon 19) or 55°C (exon 21), 30 sec at 72°C and incubation for 10 min at 72°C. PCR products were resolved by agarose gel electrophoresis to confirm successful amplification. The biotinylated products were then immobilized to streptavidin-coated beads using a solution from a commercial PSQTM96 sample preparation kit. Beads (3 μl) were diluted in binding buffer with 10 μl biotinylated PCR products, incubated for 10 min at room temperature and then transferred to a filter probe where the liquid was removed by vacuum filtration. DNA in the denaturation solution was separated; the templates were washed with washing buffer, transferred to a PSQ 96 SNP plate and annealed with the following sequencing primers: exon 18 E709K, 5′TGATCTTTTTGAATTCAGTT-3′ and G719A, 5′-CCGAAC GCACCGGAG-3′; exon 19 deletion, 5′-ATTCCCGTC GCTATC-3′; exon 20 T790M, 5′-GATGCCCAGCAGGCG-3′; and exon 21 L858R and A859T, 5′-AAGATCACAGAT TTTGG-3′ in annealing buffer at room temperature. Finally, the samples were analyzed using a PyroMark ID System and SNP reagent kit (both purchased from Biotage, Uppsala, Sweden).

The association between the EGFR mutational status and clinical characteristics, including gender, smoking status, stage, tumor specimen, pathological subtype and clinical response to EGFR-TKI, was assessed using the χ² test. The age between the two groups was analyzed by the Mann-Whitney U test. Using binary logistic regression, the EGFR mutations, according to gender, pathological subtype and smoking status, were evaluated to investigate the effect of each variable on the development of the EGFR mutations. All analyses were performed using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). P<0.05 was considered to indicate a statistically significant difference.

In total, 202 tumor samples were tested for EGFR mutations in exons 18, 19, 20 and 21 (Table I). The median age of the enrolled patients was 64 years (range, 25–87 years). Overall, 71 of the patients were female (35.1%) and 69 were non-smokers (34.2%). The majority of the histological diagnoses examined were adenocarcinomas (71.8%), followed by squamous cell carcinomas (SCCs; 19.3%; Table II). EGFR mutations were more frequent in female (68.5%), non-smoking (61.1%) and adenocarcinoma patients (94.4%). Unexpectedly, cytological were superior to biopsy specimens in determining the EGFR mutation status (45.2 vs. 21.9 %, P=0.02), indicating that a few well-selected malignant cells were sufficient to analyze EGFR mutations.

Mutations were identified in 54 (26.7%) of the 202 patients by pyrosequencing (Table III). Of the 55 mutations, the most common mutations were in-frame deletions at exon 19 and an arginine-for-leucine substitution at amino acid 858 (L858R); the frequencies of these mutations were 35 (63.6%) and 17 (30.9%), respectively. Each of the following mutations was identified once; glycine-for-alanine substitution at amino acid 719 (G719A) in exon 18, threonine-for-methionine substitution at amino acid 790 (T790M) in exon 20 and alanine-for-threonine substitution at amino acid 859 (A859T) in exon 21. Notably, one patient had in-frame deletions in exon 19 and T790M in exon 20. However, the patient did not receive EGFR-TKI treatment and therefore response to targeted drug treatment was not evaluated.

The positive predictive value (PPV) of each clinical factor was analyzed (Table IV). Among 71 females, mutations were detected in 37 patients (52.1%) and the PPVs of the non-smoker and adenocarcinoma patients were 47.8% (33/69) and 40.8% (51/125), respectively. Although EGFR mutations are commonly detected in female patients with adenocarcinoma who had never smoked, an accurate estimate of the mutation status solely from clinicopathological parameters was not obtained. The proportions of patients who had one, two or three favorable clinicopathological factors out of 54 patients who had an EGFR mutation were 27.8 (15/54), 20.4 (11/54) and 51.9% (28/54), respectively. No male patients with a non-adenocarcinoma histology who previously smoked had an EGFR mutation.

Binary logistic regression analyses were conducted to predict the EGFR criteria according to gender, pathological subtype (i.e., adenocarcinoma) and smoking status. The results are presented in Table V. Gender and pathological subtype were important variables with respect to the EGFR mutations. Gender was identified to be significantly correlated with EGFR genotype (P=0.002) and the odds ratios indicated that females had an increased probability of having EGFR mutations. Similarly, a significant correlation was identified between histological results and mutation status (P=0.007). Patients with adenocarcinoma was more closely associated with mutation status compared with other pathologies. However, by logistic analyses, smoking status was not identified to have a significant effect on EGFR mutation status in the samples. The Wald value (0.17) indicated an insignificant difference of 0.683, which revealed that non-smoking was unlikely to be directly associated with mutation status.

Of 202 patients who were tested for EGFR mutations, 92 patients were treated with EGFR-TKIs in their treatment course. Among 38 patients with positive mutations detected by pyrosequencing, 2 (5.3%) had a complete response, 26 (68.4%) had a partial response and 4 (10.5%) had stable disease. The best patient responses to EGFR-TKI treatment yielded a total disease control rate of 94.1% (Table VI). Mutation-positive patients demonstrated a more favorable prognosis and longer progression-free survival (mutation-positive vs. negative patients, 9.2 vs. 2.6 months; P<0.001). This observation was as expected. The present results indicate that EGFR mutation status identified by pyrosequencing is well correlated with the response to EGFR-TKI treatment of lung cancer patients.