Correlation between development of rash and efficacy of EGFR inhibitors: A meta-regression analysis.

Abstract

e13539 Background: Treatment with Epidermal Growth Factor (EGFR) inhibitors is frequently characterized by the presence of a rash. Previous research has suggested that developing this rash may be associated with improved outcomes. It has been difficult, however, to establish a direct link between occurrence of a rash and tumor response. Rashes are also likely to occur with greater drug exposure and it is possible that they exist as a confounder in the well established relationship between drug exposure/dosing and tumor response. The purpose of this meta-regression analysis is to define whether a correlation exists between the incidence of rash and clinical outcomes including response rate (RR), progression-free survival (PFS) and overall survival (OS) and then examine whether a similar relationship exists with other adverse events as that would suggest that rash occurrence is simply a function of drug dosing and not necessarily an indicator of tumor response. Methods: A literature search for RCT’s involving EGFR inhibitors in non-small cell lung cancer (NSCLC) identified 31 trials. This included 5 trials of Cetuximab, 13 trials of Gefitinib and 13 trials of Erlotinib. Multivariate meta-regression analysis was performed to evaluate whether all rashes (grades 1-5) and severe rashes (grades 3-5) as well as vomiting and neutropenia (two randomly chosen adverse events) may have influenced effect size (measured by RR, PFS and OS) across this cancer type. Results: The presence of severe rashes in the treatment arm was positively correlated with increased PFS (p<0.01) in NSCLC - a similar correlation was also found for risk of vomiting. Additionally, the presence of any rash in the treatment arm was positively correlated with increased OS (p=0.03) in NSCLC – a similar correlation was found for risk of vomiting and neutropenia. Conclusions: The presence of a rash seems to connote an increased likelihood of tumor response and/or survival in NSCLC. The fact that other adverse events (vomiting and neutropenia) are more evident in these responding patients suggests that rash occurrence may simply be a confounder in the EGFR inhibitors’ dose-response relationship as opposed to being independently associated with any tumor response.