Biomarker analyses from an open-label study of erlotinib in advanced non-small-cell lung cancer (NSCLC).

19012 Background: The global open-label TRUST study of erlotinib in NSCLC provides a valuable opportunity to assess biomarker hypotheses in a large patient (pt) population. A biomarker analysis of German TRUST pts was reported at ASCO 2007. Here we report data from non-German centers.,Eligibility criteria: pts with stage IIIB/IV NSCLC who had previously failed on or were unsuitable for chemo-/radiotherapy. Erlotinib 150 mg/d p.o. was given until disease progression or unacceptable toxicity. Retrospectively-obtained tumor samples were analyzed by immunohistochemistry (IHC; EGFR expression), fluorescent in-situ hybridization (FISH; EGFR gene copy number) and DNA sequencing (EGFR, KRAS gene mutations).,171 pts were included in this analysis. Baseline characteristics (%): male/female 60/40; Caucasian/other 71/29; adenocarcinoma/squamous-cell/other 56/27/17; never-smoker/ever-smoker 30/70; erlotinib 1st/2nd/3rd-line 9/53/38. Of 155 pts with response data, 13% had CR/PR and 57% had SD; disease control rate = 70%. Median progression-free survival (PFS) was 17.3 wks. EGFR IHC+ pts (≥10% membrane staining; 99/135; 73%) tended to have longer overall survival (OS) and PFS than IHC- pts (36/135; 27%); HR 0.73 in both cases, not significant (ns). OS tended to be greater in EGFR FISH+ (Cappuzzo scoring; 32/56; 57%) vs FISH- (24/56; 43%) pts (HR 0.75; ns), but PFS was similar in these two groups (HR 0.91; ns). OS and PFS were significantly greater in pts with EGFR mutations (11/75; 15%) vs pts with wild-type EGFR (64/75; 85%); HR 0.32 (p=0.013) and 0.43 (p=0.016), respectively. All pts with EGFR mutations achieved PR/SD; the response rate was significantly higher in pts with mutations (46% vs 7% for wild-type; p=0.00043). Pts with KRAS mutations (14/92; 15%) had shorter OS and PFS vs wild-type KRAS (78/92; 85%); HR 1.73 and 1.37, respectively (ns). Interestingly, 3 pts with KRAS mutations had PR.,Pts with EGFR mutation+, EGFR IHC+ or EGFR FISH+ tumors tend to have better outcomes on erlotinib therapy. However, the single-arm design of TRUST precludes conclusions on the predictive vs prognostic value of these biomarkers, and their predictive value is being assessed prospectively in a randomized phase III study of erlotinib (SATURN). [Table: see text].