LBA7509 Background: This phase III study (V-15–32) compared gefitinib vs docetaxel on overall survival (OS) in Japanese patients (pts) with pretreated advanced NSCLC. Methods: Pts with advanced or metastatic (Stage IIIb or IV) NSCLC who failed 1 or 2 chemotherapy regimens were randomized to gefitinib (250 mg/day) or docetaxel (60 mg/m2 every 3 weeks). Non-inferiority of the primary endpoint, OS, was assessed by the confidence interval (CI) of the hazard ratio (HR; gefitinib/docetaxel) derived from an unadjusted Cox proportional hazard model. Results: 489 eligible pts were recruited. Non-inferiority in OS was not achieved (HR 1.12; 95.24% CI 0.89, 1.40) according to predefined criterion (upper CI limit for HR <1.25); however, no significant difference in OS (p=0.330) or PFS (p=0.335) was apparent between treatments. Post study, 36% of gefitinib-treated pts received subsequent docetaxel and 40% received no other therapy apart from gefitinib; 53% of docetaxel-treated pts received subsequent gefitinib and 26% received no other therapy apart from docetaxel. Gefitinib significantly improved ORR (22.5% vs 12.8%; p=0.009), TTF (HR 0.63; 95% CI 0.51, 0.77; p<0.001), and QoL (FACT-L trial outcome index 20.5% vs 8.7%; p=0.002; FACT-L 23.4% vs 13.9%; p=0.023), vs docetaxel. Additional subgroup analyses will be presented. Grade 3/4 AEs occurred in 40.6% (gefitinib) and 81.6% (docetaxel) of pts. Incidence of interstitial lung disease (ILD) was 5.7% (n=14) and 2.9% (n=7), respectively. There were 4 deaths due to AEs in the gefitinib arm (3 possibly treatment-related due to ILD; 1 due to pneumonia that was not considered treatment-related) and none in the docetaxel arm. Conclusions: Whilst non-inferiority in OS between gefitinib and docetaxel was not demonstrated according to predefined criteria, there was no statistically significant difference in survival between the two arms. Secondary endpoints largely unaffected by subsequent therapy provide further evidence of clinical efficacy of gefitinib in these pts. AEs were consistent with those previously observed for both treatments. [Table: see text]