Frequent epidermal growth factor receptor gene mutations in malignant pleural effusion of lung adenocarcinoma

Patients and tissue procurement

Between June 2005 and December 2006, 383 consecutive samples of pleural effusion were collected from 273 patients who had received thoracentesis in the chest ultrasonography examination room of the National Taiwan University Hospital (Taipei, Taiwan). This study was approved by the institutional review board of the National Taiwan University Hospital, and all patients had signed an informed consent form before the thoracentesis was performed. The primary malignant tumours or MPEs were confirmed by pathology or cytology reports. Of the 273 patients, 164 had lung adenocarcinoma and the other 109 patients had malignancy other than lung adenocarcinoma or nonmalignancy-related pleural effusions. Among the 164 patients with lung adenocarcinoma, the cytology of pleural effusions was negative in 28 patients. Therefore, 136 MPEs of lung adenocarcinoma were assessed for EGFR mutations. These 136 MPEs were obtained before treatment with gefitinib, or erlotinib if the patients received EGFR TKI therapy. For comparison, 91 surgically resected specimens of lung adenocarcinoma archived from April 2001 to November 2004 were also retrived. Informed consent about the use of these specimens for future molecular studies with approval of the institutional review board was obtained before surgery. The retrospective use of archival tissue for EGFR gene analysis was approved by the institutional review board of the National Taiwan University Hospital. The clinical information of all the patients was recorded, including age, sex, smoking history, lung cancer stage, performance status (Eastern Cooperative Oncology Group performance status (ECOG PS)), treatment regimens and maximal response. Patients who had smoked <100 cigarettes in their lifetime were categorised as never smokers. Those who smoked cigarettes within 1 yr of the diagnosis were categorised as current smokers. The others were categorised as former smokers.

Collection of pleural effusion fluid and surgically resected specimens

The pleural effusion fluid was collected into heparinised tubes. A 1-mL sample of the fluid was centrifuged at 250×g for 10 min at room temperature, and the cell pellet frozen. Specimens of lung adenocarcinoma tissue obtained at surgery were immediately snap-frozen in liquid nitrogen and stored for later use. Total RNA was isolated using Tri-reagent (Molecular Research Center, Inc., Cincinnati, OH, USA) and stored at -80°C until RNA extraction was performed. RNA was extracted from cell lysates with a Qiamp RNA Mini Kit (Qiagen, Hilden, Germany) according to the protocol in the manufacturer's instructions. The RNA obtained was eluted in 50 μL of sterile bi-distilled buffer. The amounts of RNA extracted were measured with spectrophotometry.

PCR amplification and direct sequencing

The four exons (exons 18–21) that code for the tyrosine kinase domain of the EGFR gene were amplified with a forward primer (5′-AGCTTGTGGAGCCTCTTACACC-3′) and reverse primer (5′-TAAAATTGATTCCAATGCCATCC-3′), as reported in a prior study 8. The RT-PCR was performed previously as reported 8, using a Qiagen OneStep RT-PCR Kit (Qiagen). The RT-PCR conditions were as previously described 8. RT-PCR amplicons were purified and sequenced using the Big Dye Terminator sequencing kit (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's instructions. Sequencing products underwent electrophoresis on an ABI PRISM 3100 (Applied Biosystems). Both the forward and reverse sequences obtained were analysed by basic local alignment search tool against the cDNA sequence of the EGFR gene ( accession number NM005228) and chromatograms were examined manually. PCR amplicons were sequenced in both sense and antisense directions. EGFR mutations detected in the initial round of sequencing were confirmed by subsequent rounds of independent RT-PCR and sequencing reactions. Only specimens in which a mutation was identified in both rounds were recorded as mutation positive. Mutations were also checked against the single nucleotide polymorphism database.

Antitumour therapy and response evaluation of the patients with MPEs of lung adenocarcinoma

The antitumour response of the patients was evaluated by chest radiography every 2–4 weeks and by computed tomography of the disease sites every 8–12 weeks after treatment; this is routine practice in the present authors’ institution. For the measurable solid tumours, the unidimension method was used according to the Response Evaluation Criteria in Solid tumours (RECIST) guidelines 18. Partial response and stable disease were confirmed by a sustained 4-week follow-up. Overall survival after antitumour therapy was calculated from the date of initiation of treatment to the date of death, last follow-up or the final follow-up day of the study, which was September 1, 2007.

Toxicity of EGFR TKIs was also recorded, and was graded according to the National Cancer Institute Common Toxicity Criteria Version 3.0.

Statistical analysis

All categorical variables were analysed with Pearson’s Chi-squared test, except where a small size required the use of Fisher's exact test. Overall survival was compared by the log-rank test, and multivariate analysis for overall survival was performed using the Cox linear regression method. Two-sided p-values <0.05 were considered significant.

Clinical characteristics and EGFR mutations of the surgically resected lung adenocarcinoma patients

In total, 91 surgically resected specimens were assessed for the EGFR mutation (table 1⇓). The median (range) age of the 91 patients was 63.4 (37.5–85.4) yrs. Three stage-IIIB patients had satellite tumour nodules within the same lobe as the primary tumour, and they received lobectomies. Two stage-IV patients, who both had two lung lesions over the right middle lobe and right lower lobe without other distant metastasis lesions, received bi-lobectomies.

In total, 46 out of 91 surgically resected lung adenocarcinoma had EGFR mutations. The EGFR mutation rate was 50.5%. The EGFR mutations included 19 L858R, 23 deletions in exon 19 and four other types (767–769 duplication ASV, P771_H773insYNP+H773Y, K860I+L861Q and L861Q). EGFR mutation rates of the surgically resected lung adenocarcinoma were not significantly different by sex (49% of males versus 52.5% of females; p = 0.742), age (p = 0.746) or smoking status (54.7% of never smokers versus 44.7% of former or current smokers; p = 0.348; table 1⇑).

Of the 91 patients, 14 took EGFR TKIs after tumour recurrence. Of these, seven patients had EGFR mutations, and the other seven patients were wild type. Five patients with a positive EGFR mutation and three patients of wild type responded to EGFR TKI treatment. However, the patient numbers were too small for further analysis.

Clinical characteristics and EGFR mutations of the lung adenocarcinoma patients with MPE

MPEs were assessed for EGFR mutations in 136 patients (table 2⇓). The median (range) age of the 136 patients was 66.1 (28.7–90.6) yrs. Two (1.5%) patients did not complete staging work-up of lung cancer and they were lost to follow-up.

In total, 93 out of 136 patients with MPEs of lung adenocarcinoma had positive EGFR mutations. The mutation rate was 68.4%. The EGFR mutations included 50 L858R, 32 deletions in exon 19, and 11 other types. The 11 other types included three G719A, one G719A+S720F, one G719A+S768I, one E746V+L747P, one L747P, one 767–769 duplication ASV, one 768–770 duplication SVD, one L861Q and one R776H+L861Q. EGFR mutation rates of the adenocarcinoma patients with MPEs were not significantly different by sex (64.2% of males versus 71.1% of females; p = 0.396), age (p = 0.734) or smoking status (70.5% of never smokers versus 63.4% of former or current smokers; p = 0.413; table 2⇑).

Of the 136 patients with MPEs, 111 patients initially had pleural effusions at the diagnosis of lung adenocarcinoma. Of the 111 patients, 77 (68.5%) had EGFR mutations. The mutation rate of initial pleural effusions related to lung adenocarcinoma was similar to the total 136 MPEs (68.5% versus 68.4%).

The patients with MPEs of lung adenocarcinoma had a higher EGFR mutation rate than the patients with surgically resected lung adenocarcinoma (68.4% versus 50.5%; p = 0.007; table 3⇓). Interestingly, the mutation rates of deletion in exon 19 in MPEs and surgically resected specimens were not significantly different (23.5% versus 25.3%). However, the patients with MPEs of lung adenocarcinoma had a higher L858R mutation rate than the patients with surgically resected lung adenocarcinoma (36.8% versus 20.9%; p = 0.011; table 3⇓).

Response of adenocarcinoma with MPEs treated with EGFR TKI

Of the 136 patients with MPEs of lung adenocarcinoma, 71 had received EGFR TKI treatment (gefitinib 250 mg·day⁻¹ or erlotinib 150 mg·day⁻¹). Two patients were lost to follow-up before treatment and the other 63 patients received other antitumour therapy. Of the 71 patients treated with EGFR TKI, 38 had partial response. Those included one wild type, 23 L858R, 11 deletions in exon 19, one G719A, one L861Q and one combined EGFR mutation (G719A+S720F; table 4⇓). For the overall survival analysis of the 136 patients with lung adenocarcinoma and MPEs, the median overall survival was longer for patients with EGFR mutations than for patients with wild-type EGFR (median 21.4 months, 95% confidence interval (CI), 17.9–24.9 months versus 11.5 months, 95% CI 7.4–15.6 months; p = 0.005 by log-rank test). Multivariate analysis was performed by the Cox regression model for the potential prognostic factors, including: age, smoking status, ECOG PS, EGFR mutation (table 5⇓). Lung adenocarcinoma patients with MPEs who were aged <65 yrs (p = 0.011), who never smoked (p = 0.027), with ECOG PS 0–1 (p<0.001) and EGFR mutation (p = 0.001) were found to be associated with a longer overall survival.

The toxicity of EGFR TKIs were distributed from grade 1 to grade 3, and no patients had grade 4 toxicity. The most common toxicity included diarrhoea (24%), skin rash (38%), acne (31%), dry skin (22%), pruritus (22%) and anorexia (20%). Grade 3 events included one diarrhea (1%), one skin rash (1%) and one acne (1%). Neither grade 4 toxicity nor interstitial pneumonitis was recorded.