LBA-01 - Survival outcomes according to body mass index (BMI): results from a pooled analysis of 5 observational or phase IV studies of bevacizumab in metast...

Date 04 July 2015
Event WorldGI 2015
Session Oral and LBA abstracts
Topics Anti-Cancer Agents & Biologic Therapy
Cancer Aetiology, Epidemiology, Prevention
Colon Cancer
Rectal Cancer
Presenter Y. Zafar
Citation Annals of Oncology (2015) 26 (suppl_4): 117-122. 10.1093/annonc/mdv262
Authors Y. Zafar1, A. Grothey1, E. Van Cutsem2, J. Hubbard1, F. Hermann3, A.J. Storm4, E. Gomez4, C. Revil4
  • 1Duke University Medical Center, Durham/
  • 2Leuven Cancer Institute, University Hospitals Leuven, Leuven/BE
  • 3University Hospitals Leuven, Leuven/BE
  • 4F. Hoffmann-La Roche Ltd. , Basel/CH

Abstract

Introduction

High BMI has been associated with increased risk of colorectal cancer. However, little is known about how BMI impacts outcomes for patients already diagnosed with mCRC. Here we describe an analysis of overall survival (OS) and progression-free survival (PFS) according to BMI in a pooled dataset of patients treated in prospective, observational studies.

Methods

Data from patients with previously untreated mCRC who received bevacizumab with chemotherapy were pooled from the prospective BEAT (ex-USA), BRiTE (USA), AWB (Germany), CONCERT (France), and ARIES (USA) studies. OS and PFS were evaluated via the Kaplan-Meier method in first-line patients stratified by baseline BMI categories. The ARIES study did not calculate patient BMI; data from ARIES were excluded from the present analyses.

Results

BMI data were available for 6128 of 7688 patients. Median BMI for all patients was 25.3 kg/m2 (IQR: 22.6–28.7). Median OS and PFS according to patients grouped by BMI are shown in the table.

Conclusion

In this pooled dataset of patients with mCRC treated with first-line bevacizumab and chemotherapy in observational/phase IV trials, patients with the lowest BMI had shorter median OS, suggesting that low BMI could be a poor-prognostic factor. PFS was similar between subgroups. Additional analyses, including adjusted Cox models supporting this finding will be presented at the meeting.