Purpose:Preclinical models suggest bidirectional signaling between estrogen receptor(ER) and epidermal growth factor receptor (EGFR) pathways in lung cancer. This phase II randomized clinical trial of erlotinib, an EGFR inhibitor, and fulvestrant, an ER antagonist, versus erlotinib alone, evaluated efficacy in advanced stage non-small cell lung cancer (NSCLC) patients. The purpose of this investigation is to evaluate the influence of various clinical, biochemical, and histopathological features as predictive or prognostic markers.Methods:Patients who met eligibility criteria for the study were enrolled in a 2:1 ratio tocombination erlotinib and fulvestrant or erlotinib alone. Primary endpoint was objective tumor response rate, while secondary analysis included overall survival (OS) and progression free survival (PFS). Analysis included blood estrogen (estradiol and estrone) run in duplicate, and a panel of relevant biomarkers obtained at the beginning of the first three 28-day cycles. Pre-treatment tumor tissue was obtained and evaluated using immunohistochemistry (IHC) for estrogen receptor (ER)-α, ER-β, and progesterone receptor.Results:106 patients were randomized to a treatment arm between March 2006 and June2011 with 100 patients receiving at least one dose of therapy. Clinical response data were available for 82 patients. Median PFS was 4.3 and 1.9 months for the erlotinib plus fulvestrant arm versus the erlotinib alone arm (HR 0.64, 95% CI 0.34-1.08). Median OS was 11.6 and 8.6 months for the erlotinib plus fulvestrant versus the erlotinib alone arm (HR 0.9876 95% CI 0.60-1.6). Blood analysis showed that after 1 cycle of treatment, placental growth factor (PlGF) and transforming growth factor-α (TGF-α) were higher in patients that had progressive disease (PD) compared to partial response (PR) plus stable disease (SD),p=0.02 and p=0.02, respectively. In the tissue, ER-α or progesterone receptor positivity did not demonstrate a significant difference in response for patients treated with the combination erlotinib and fulvestrant, Chi-squared test, p=0.96. Moreover, cytoplasmic ER-β H score was higher in patients with PD versus PR + SD (median 90 versus 20 respectively, p=0.01). The dataset did not demonstrate statistical significance by treatment arm (median 60 versus 65 in the erlotinib versus erlotinib plus fulvestrant, p=0.59).Conclusion:Combination erlotinib and fulvestrant in this study did not show superiority overerlotinib alone with respect to OS and PFS. Blood estrogen and estrone levels did not correlate with response or treatment arm. Biomarkers that may be prognostic include cytoplasmic ER-β H score, while higher Cycle 2 Day 1 PlGF and TGF-α levels may suggest a poor response.