Premium
Bevacizumab exposure beyond first disease progression in patients with metastatic colorectal cancer: analyses of the ARIES observational cohort study
Author(s) -
Grothey Axel,
Flick E. Dawn,
Cohn Allen L.,
BekaiiSaab Tanios S.,
Bendell Johanna C.,
Kozloff Mark,
Roach Nancy,
Mun Yong,
Fish Susan,
Hurwitz Herbert I.
Publication year - 2014
Publication title -
pharmacoepidemiology and drug safety
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.023
H-Index - 96
eISSN - 1099-1557
pISSN - 1053-8569
DOI - 10.1002/pds.3633
Subject(s) - bevacizumab , medicine , hazard ratio , proportional hazards model , colorectal cancer , oncology , confounding , confidence interval , adverse effect , cohort study , cohort , surgery , cancer , chemotherapy
Purpose This analysis from Avastin® Registries: Investigation of Effectiveness and Safety (ARIES) examined the association between exposure to bevacizumab after disease progression (PD) and postprogression survival (PPS) in bevacizumab‐exposed metastatic colorectal cancer (mCRC) through the application of time‐dependent and time‐fixed analytical methods. Methods Patients with mCRC who were treated with first‐line bevacizumab and who survived first PD (PD1) were included. A time‐dependent Cox regression model was fitted to assess the effect of cumulative bevacizumab exposure on PPS, while controlling for potential confounders. In addition to support findings from previous studies, a modified intent‐to‐treat (mITT) analysis compared PPS in patients who received bevacizumab beyond disease progression (BBP) with those who did not (No‐BBP). Results Of 1550 patients, 1199 survived PD1 and had a median PPS of 13.4 months. Cumulative bevacizumab exposure was associated with improved PPS ( p = 0.0040). After adjusting for confounders, the hazard ratios (HRs) for PPS decreased, on average, by 1.2% (range, 1.1–1.3%) with each additional dose of bevacizumab. In the mITT analysis, the median PPS for BBP ( n = 438) was 14.4 months vs 10.6 months with for No‐BBP ( n = 667). BBP was found to be independently associated with longer PPS in a multivariable Cox regression analysis (HR, 0.84; 95% confidence interval, 0.73–0.97). Protocol‐specified adverse events suspected to be associated with bevacizumab occurred in 13.0% of patients with BBP. Conclusion This analysis supports the observation that bevacizumab exposure after PD1 is associated with longer PPS in mCRC. Copyright © 2014 John Wiley & Sons, Ltd.