7587 Background: PIK3CA encodes the p100a subunit of the mitogenic signaling protein phosphatidylinositol 3-kinase (PI3K). PI3K is a RAS-activated tyrosine kinase that signals through downstream mediators including AKT and m-TOR. PIK3CA mutations in the helical binding domain and catalytic subunit are implicated in tumorigenesis and treatment resistance in many cancers. The clinical characteristics of patients with PIK3CA-mutant lung cancer have not been reported. Methods: To identify patients with NSCLC harboring PIK3CA mutations, we examined results of routine assessment of histology and driver mutations in EGFR, KRAS, BRAF, HER2, PIK3CA, AKT1, NRAS, MEK1, and EML4-ALK (Kris, Proc ASCO, 2010). Clinical data were extracted from the medical records of patients with PIK3CA mutant tumors. Results: 25 of 1325 (2%, 95% CI 1-3%) patients had a mutation in PIK3CA (11 E545K, 3 E542K, 3 H1047L, 8 H1047R). The patients (64% women) had a median age of 66 (range 34-78). 9 patients (36%, 95% CI 20-56%) were never-smokers. Tumor histologies included: 23 adenocarcinomas, 1 squamous, and 1 large cell neuroendocrine. Stages at presentation were: 10, IV; 2, IIIB; 6, IIIA; 2, IIA; 1, IB; and 4, IA. The median survival of patients with stage IIIB and IV was 21 months. 16 of 25 (64%, 95% CI 44-80%) had co-existing mutations in other oncogenes: 10, KRAS; 1, MEK1; 1, BRAF; 1, EML4-ALK; and 3, EGFR exon 19 deletions. Of the 3 patients with EGFR mutations, only 2 tumors responded to erlotinib, with acquired resistance developing at 5 and 15 months. The tumor with BRAF responded to gefitinib. Stable disease was the response in the patient with EML4-ALK given crizotinib. Conclusions: We found PIK3CA mutations in all histologic types of non-small cell lung cancer. They are not mutually exclusive with mutations in EGFR, KRAS, BRAF, and EML4-ALK. Data on the impact of PIK3CA mutations on the efficacy of targeted therapies (erlotinib, crizotinib) are scant. The testing of agents targeting PIK3CA requires complete genotyping of tumor specimens to assess the potentially confounding role of co-existing mutations to affect responses. Support: NIH T32, P01 CA129243, RC2 CA148394.