1. ¹Hyogo Medical Center for Adults, Respiratory Medicine, Akashi, Japan
2. ²Hyogo Medical Center for Adults, Pathology, Akashi, Japan
3. ³Hyogo Medical Center for Adults, Radiology, Akashi, Japan
4. ⁴Aichi Cancer Center Hospital, Pathology and Molecular Diagnosis, Nagoya, Japan
5. ⁵Aichi Cancer Center Hospital, Thoracic Surgery, Nagoya, Japan

Correspondence: Dr M Satouchi, Hyogo Medical Center for Adults, Respiratory Medicine, 13-70 Kitaoji-cho, Akashi, Hyogo, 673-8558, Japan; E-mail: satouchi@hp.pref.hyogo.jp

Received 4 December 2006; Revised 28 February 2007; Accepted 28 February 2007; Published online 27 March 2007.

Abstract

This study aimed to identify predictive factors associated with prognostic benefits of gefitinib. A total of 221 Japanese patients who received gefitinib (250 mg day^-1) were examined retrospectively and potential predictive factors analysed. Overall response rate (ORR) was 24.4% and median survival time (MST) was 8.0 months. In a log-rank test, survival was significantly better in females, patients with adenocarcinoma, never-smokers, favourable performance status (PS) and patients with epidermal growth factor receptor (EGFR) mutation. The lower the smoking exposure (Brinkman Index (BI)=cigarettes per day years smoked), the better the MST (BI 0: 14.5 months, BI <500: 9.5 months, BI 500 to <1000: 6.9 months, BI 1000: 4.0 months). Positive-EGFR mutation status and PS 0–1 were independent predictors of favourable prognosis by multivariate analysis. Prognosis was significantly different according to EGFR mutation status (with the same smoking status), but not according to smoking status (with the same EGFR mutation status). EGFR mutation status is the most important independent predictor of survival benefit with gefitinib treatment. Although differences in prognosis were observed according to relative smoking status and smoking exposure, the results suggested that smoking is not a direct predictor of prognosis, yet is a surrogate marker of EGFR mutation status.