7013 Background: In two Phase III trials (INTACT 1 & 2), gefitinib (‘Iressa’, ZD1839) in combination with first-line platinum-based chemotherapy regimens failed to show benefit in terms of survival when compared with chemotherapy alone in patients (pts) with NSCLC. The primary objective of this analysis was to evaluate if EGFR staining was predictive of survival. Methods: Pretreatment tumor biopsies were assessed by immunohistochemistry for EGFR using the DakoCytomation pharmDx assay™. The percentage of tumor cells was assessed using 4 levels of intensity: no staining, 0; weak, 1+; moderate, 2+; strong, 3+; as well as the presence of complete membrane staining. 516 pts (INTACT 1/2, 219/297) with tumor samples fully evaluable for EGFR were analyzed in the per protocol population of the combined trials. The analysis was stratified by trial and performed independently for pts randomized to placebo or gefitinib. A restricted backwards elimination Cox regression analysis was conducted to identify independent EGFR factors. Variables found to be significant (p<0.1) were also tested for treatment interaction to determine if they served as predictive factors. Results: The demographics of the analysis population were representative of the overall population in each trial as were common disease characteristics. The analysis showed that two EGFR-based variables are independent strong prognostic factors for both placebo- and gefitinib-treated patients. However, no interactions with treatment (ie no value for predicting benefit of gefitinib treatment) were indicated for either of these two variables (p=0.8688 and 0.4967) or for any other EGFR-based variables. Conclusion: Given the overall trial results for INTACT 1 & 2 showing no survival benefit between gefitinib and placebo, it was highly unlikely that EGFR variables would identify a subset of pts who receive survival benefit. The data presented support this hypothesis. 'Iressa' is a trademark of the AstraZeneca group of companies 'PharmDx assay' is a trademark of DakoCytomation Author Disclosure Employment or Leadership Consultant or Advisory Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration AstraZeneca Pharmaceuticals; DakoCytomation AstraZeneca; DakoCytomation; Genentech AstraZeneca AstraZeneca AstraZeneca Pharmaceuticals

7024 Background: Two double-blind placebo-controlled randomized multicenter trials (INTACT 1 & 2) in NSCLC failed to show survival benefit for gefitinib (‘Iressa’, ZD1839) in combination with first-line platinum-based chemotherapy. However, these large studies provide an opportunity to evaluate if EGFR staining is prognostic for survival. Methods: Tumor biopsies (n=516) taken between diagnosis and start of therapy were assessed by immunohistochemistry for EGFR using the DakoCytomation EGFR pharmDx assay™. The analysis was stratified by trial and performed independently for patients randomized to placebo and gefitinib. A restricted backwards elimination Cox regression analysis was conducted to identify independent EGFR factors that were statistically significant (p<0.10) and these were also tested for treatment interaction to assess if they served as predictive factors. Results: The analysis found 2 statistically significant EGFR-based prognostic factors representing the growth pattern and percent membrane staining in patients treated with gefitinib (p=0.0023, df=2) and placebo (p=0.0128, df=2). A further analysis including interaction terms of the EGFR variables with treatment confirmed these as highly significant (p<0.0001, df=2) but not the interaction. The EGFR factors remained independently prognostic when conditioning for previously found factors (PS2; stage IV; bone, brain, liver metastasis; sex; histology; weight loss) and were comparably significant (p=0.0011 and p=0.0158). Results will also be presented for secondary endpoints (time to progression, objective response). Conclusion: While the majority of previous studies indicate higher EGFR expression correlates with poor survival, the analysis from 2 large Phase III trials provides statistically significant evidence that the combination of EGFR expression and growth pattern is a strong prognostic indicator for improved survival within this setting. ‘Iressa’ is a trademark of the AstraZeneca group of companies ‘PharmDx assay’ is a trademark of DakoCytomation Author Disclosure Employment or Leadership Consultant or Advisory Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration AstraZeneca Pharmaceuticals; DakoCytomation DakoCytomation; Genentech AstraZeneca AstraZeneca