Flynn, P.

3537 Background: Bevacizumab (BV) prolongs overall survival and progression-free survival (PFS) when added to standard chemotherapy (CT) in patients with metastatic colorectal cancer (mCRC). BRiTE is a large, community-based observational registry of patients with mCRC receiving BV plus first-line CT. Safety and efficacy information in unselected patients with mCRC are collected. Chemotherapy regimen choice is at the physician’s discretion. Methods: 1968 patients receiving BV plus first-line CT were enrolled at 248 sites in 49 states between Feb 2004 and Jun 2005. Patients are followed for up to 3 years, with data reporting every 3 months (mo) including disease status as assessed by investigator using his/her method of choice. Patients were grouped twice by CT. Cox regression models, adjusted and unadjusted for treatment assignment, were used to identify baseline characteristics (BC) and differential chemotherapy effects on PFS. Results: Cohort demographics were consistent with the NCI Surveillance, Epidemiology, and End Results (SEER) database for mCRC. Median study follow-up was 10 months by Nov 4, 2005. CT regimens included FOLFOX (55.8%), FOLFIRI (14.1%), and IFL (9.7%). Based on observed progression events (n=800) or deaths (n=123), projected median PFS is 11.3 mo (95% CI: 10.4–11.7). Median PFS was comparable in patients treated with CT regimen based on irinotecan (11.3 mo), oxaliplatin (11.4 mo), or neither (10.2 mo) (CT Grouping 1); or 5FU infusion (11.5 mo), 5-FU bolus (9.7 mo), or capecitabine (11.6 mo) (CT Grouping 2). Significant BC in adjusted and unadjusted models are ECOG performance status (PS), primary disease site, and albumin, with no significant difference among CT regimens. Conclusions: This large, community-based registry included a patient population with demographics consistent with the SEER database for mCRC. Preliminary median PFS of 11.3 mo compares favorably to that reported in the pivotal trial. Median PFS appears to be comparable for all CTs when combined with BV. Higher baseline albumin, PS of 0, and primary disease site of rectum were associated with improved outcome. [Table: see text]

1071 Background: N0234 is a phase II trial of gemcitabine at 1000 mg/m2 days 1 and 8 of a 21 day cycle and erlotinib at 150 mg daily in patients with previous therapy with either an anthracycline or a taxane for MBC. A partial response (PR) rate of 17% has been reported (ASCO 2005). Recent information suggests that over-expression of the EGFR pathway is frequent in patients with TNeg MBC. Methods: Tissue and serum were obtained for translational studies of ER, PR, HER2, and proteins related to EGFR pathway on tissue using immunohistochemistry; CK19, mammoglobin, and TTK on circulating tumor cells (CTCs). Markers were evaluated centrally, blinded to patient outcome. Complete data were available in 57 of 59 eligible patients; 20/57 patients (pts) were TNeg. Results: A PR rate of 25% and clinical benefit rate (PR + stable disease > 6 months)of 25% was observed in TNeg pts as compared to a PR rate of 14% (p = 0.30) and clinical benefit rate 22% (p= 0.75) for non-TNeg pts. Median progression free survival 72 days vs 98 days (p=0.13) and overall survival 227 days vs 738 days (p<0.001) were shorter for TNeg pts. Translational results for the EGFR pathway proteins and CTC analysis is forthcoming. Conclusions: The combination of gemcitabine and erlotinib appears to be as active in the non-TNeg and TNeg subset of patients on this trial. TNeg patients had an inferior clinical result, confirming the poor outcome seen in these patients. Data for the remaining biomarkers potentially predictive of activity will be presented. Supported by CA-25224, CA-37404. [Table: see text]