7575 Background: Most non-small cell lung cancers (NSCLC) are primarily resistant to the EGFR tyrosine kinase inhibitor gefitinib. One strategy to overcome resistance is to block parallel or downstream effectors that maintain tumor growth and proliferation. Preclinical data supports this hypothesis as inhibition of the PIK3CA/Akt/mTOR pathway restores gefitinib sensitivity in resistant cells. In a phase I study, gefitinib (250 mg) and the mTOR inhibitor everolimus (5 mg) were safely administered orally once-a-day (Milton Proc ASCO 2005). This combination induced regressions in patients with NSCLC. To assess the efficacy of this regimen, we designed this phase II trial. Methods: Two cohorts of patients with measurable stage IIIB/IV NSCLC were entered: (1) untreated with chemotherapy and (2) previously treated with cisplatin or carboplatin and docetaxel. A Simon two-stage design proposed to enroll up to 31 patients in each cohort. Response was assessed after 1 month then every 2 months. Pretreatment tumor specimens were collected on all patients. Results: To date, 25 patients have entered, 11 untreated and 14 previously treated. The median age was 66 years and KPS 80%. 52% were women, 68% had adenocarcinoma, and all were current (n=1) or former smokers. Partial responses (RECIST) have been observed in 4 of 23 evaluable patients (17%, 95% CI 5 to 39%) One of these 4 tumors harbored a KRAS mutation and one an EGFR exon 19 deletion. Responses lasted 3, 4, 9+, and 16+ months. Responses were observed in 2 of 13 untreated and 2 of 10 previously treated individuals. Toxicities: Grade 2/3 diarrhea in 13%, Grade 2 pustular rash in 25%, and Grade 2 mucosal ulcerations in 38%. 8% had dose reductions. There were no Grade 4 toxicities or treatment-related deaths. Conclusions: The combination of everolimus and gefitinib produced a 17% partial response rate in smokers with NSCLC, a group less likely to benefit from gefitinib. Enough responses have been seen in both untreated and previously treated cohorts to complete enrollment according to the two stage design of this trial. We plan to correlate outcomes with the molecular profile of tumors. Supported by Novartis, USA No significant financial relationships to disclose.
Approaches to handling data when a phase II trial deviates from the pre-specified Simon's two-stage design
