Purpose: It has been reported that the mutations of epidermal growth factor receptor (EGFR) are detected in lung cancers. Studies of EGFR mutations in large numbers of patients' tumors with clinical data including response to EGFR tyrosine kinase directed therapy are needed to develop a robust database for clinical use. The purpose of the present study is to gain further insights into the significance of EGFR mutation in non^small cell lung cancer (NSCLC). Experimental Design: We investigated the clinicopathologic significance of tyrosine kinase domain (exons 18-21) EGFR mutations in 120 patients with primary NSCLC and the correlation between EGFR mutation and sensitivity to gefitinib in an additional 20 NSCLC patients treated with gefitinib. In addition, onocogenic KRAS mutations and RASSF1A promoter methylation were determined in the same samples. Results: EGFR mutation was detected in 29 of 120 (24%) tumors. All of the 29 (40%) mutations occurred in 72 adenocarcinomas. EGFR mutation was significantly more frequent in females (47%) than males (12%, P < 0.0001), in younger patients (38%) than older patients (10%, P = 0.0005), in nonsmokers (47%) than smokers (13%, P < 0.0001), and in well-differentiated tumors (39%) than moderately and poorly differentiated tumors (7%, P < 0.0001). Mutation of the EGFR gene was preferentially observed in advanced disease. Furthermore, EGFR mutations were detected in 11 of 14 (79%) responders, whereas none of six (0%) nonresponders had the mutation (P = 0.0022). Conclusions: These results in Japanese (East Asian) patients indicated that EGFR mutation plays an important role in pathogenesis of lung adenocarcinoma.

The histologic prognostic factors of pulmonary adenocarcinomas of the lung less than 2 cm in diameter were analyzed in 75 patients who had undergone surgical resection. The pathologic stage, lymph node involvement, and pleural involvement were found to be the major determinants of prognosis (P less than 0.01). In addition, other single factors, such as tumor differentiation (P less than 0.01), vascular invasion (P less than 0.01), the degree of collagenization in the fibrotic focus (P less than 0.01), the standard deviation (SD) of nuclear areas (P less than 0.05), and mitotic index (P less than 0.05) correlated significantly with prognosis by the log-rank test on the Kaplan-Meier survival curves of these factors. Patients with dense infiltration of "T-zone histiocytes" survived significantly longer than those with less infiltration (P less than 0.05). Cox's proportional hazard general linear model analysis showed the importance of factors, such as lymph node or pleural involvement and the SD of nuclear area, when the pathologic stage was excluded, and of the mitotic index when all four factors were excluded to emphasize the cellular characteristics. It is possible to predict the postoperative prognosis of patients with small pulmonary adenocarcinoma more precisely by combination of the above histopathologic factors.

Gefitinib should not be given to patients with PF apparent on chest radiograph. Patients with PF on chest CT but not detected on chest radiograph could be treated carefully with gefitinib, but a risk-benefit analysis should be considered.

Intrapericardial instillation of 2 mg of mitomycin C was feasible and demonstrated a promising response against malignant pericardial effusion resulting from non-small cell lung cancer.