Previous bevacizumab and efficacy of later anti-epidermal growth factor receptor antibodies in metastatic colorectal cancer: results from a large international registry.

Abstract

BACKGROUND: The FIRE-3 [5-fluorouracil, folinic acid, and irinotecan (FOLFIRI) plus cetuximab versus FOLFIRI plus bevacizumab in first line treatment colorectal cancer (CRC)] study reported that first-line FOLFIRI plus cetuximab versus FOLFIRI plus bevacizumab resulted in similar progression-free survival (PFS) but improved overall survival (OS). A potential explanation is that the initial biologic agent administered in metastatic CRC (mCRC) affects later line efficacy of the other treatments. We sought to test this hypothesis.

MATERIALS AND METHODS: We interrogated our mCRC registry (Treatment of Recurrent and Advanced Colorectal Cancer) regarding treatment and outcome data for RAS wild-type patients receiving epidermal growth factor receptor inhibitors (EGFRIs) in second and subsequent lines. Survival outcomes from the beginning of EGFRI use were determined as a function of previous bevacizumab use and the interval between ceasing bevacizumab and beginning EGFRI use.

RESULTS: Of 2061 patients, 222 eligible patients were identified, of whom 170 (77%) had received previous bevacizumab and 52 (23%) had not. PFS and OS from the start of EGFRIs did not differ by previous bevacizumab use (3.8 vs. 4.2 months; hazard ratio [HR], 1.12; P = .81; 9.0 vs. 9.2 months; HR, 1.19; P = .48, respectively) for the whole cohort or when analyzed by the primary tumor side (HR for left side, 1.07; P = .57; HR for right side, 1.2; P = .52). PFS was significantly shorter with right-sided primary tumors when the interval between bevacizumab and EGFRI use was < 6 versus > 6 months (median, 2.2 vs. 6 months; HR, 2.23; P = .01) but not with left-sided tumors (median, 4.2 vs. 5.5 months; HR, 1.12; P = .26).

CONCLUSION: Previous bevacizumab use had no effect on the activity of subsequent EGFRIs. The apparent effect of time between biologic agents in right-sided tumors might reflect patient selection.