7521 Background: WJTOG3405 met its primary endpoint of progression free survival (PFS) (9.2 months (mo.) for G vs. 6.3 mo. for CD, hazard ratio (HR) 0.489, 95% confidence interval (CI): 0.336-0.710). (Mitsudomi et al., Lancet Oncol., 2010). However, the impact on overall survival (OS) was not clear then because of relatively short follow-up period. Methods: Overall survival (OS) was re-evaluated using updated data (data cutoff, 31 July, 2011, median follow-up, 34 months) for 172 patients. Results: Eighty-two events had occurred (48%). Median survival time (MST) for G arm was 36 mo. (95% CI: 26.3 -) which was not significantly different from 39 mo. (95% CI: 31.2 -) for CD arm (HR 1.185, 95% CI 0.767-1.829). Multivariate analysis using Cox proportional hazards model revealed that none of covariates (treatment arm, smoking status, sex, age, postoperative recurrence or IIIB/IV, and mutation type) significantly affected OS. In the G arm, MST of patients with exon 19 deletion (36 mo.) was comparable to that of patients with L858R (35 mo.). In the CD arm, 78 patients (91%) received EGFR-TKI as the 2nd or later line treatment, whereas in the G arm, 52 patients (61%) received platinum doublet. Accordingly, 130 patients received both platinum doublet and EGFR-tyrosine kinase inhibitor (TKI) and 34 patients received EGFR-TKI without platinum doublet in their whole courses of therapy. MST for the former and the latter group were 36 months (95% CI: 31.2-45.7) and 45 months (95% CI: 25.6-), without significant difference. Conclusions: This update OS analysis revealed that G for advanced NSCLC with EGFR mutation offers distinct survival benefit of 3 years. There was no difference in OS whether the first-line treatment was G or CD, in accordance with the precedent studies. The reason why PFS difference was not translated into OS difference is probably due to high cross over rate to EGFR-TKI. However, it was noteworthy that 40% of patients in the G arm could be managed without platinum doublet and yet had similar outcome.