437TiP - A randomised phase II study of carboplatin and bevacizumab in recurrent glioblastoma multiforme (CABARET study). Cooperative trials Group for Neuro-...

Date 30 September 2012
Event ESMO Congress 2012
Session Poster presentation II
Topics Cytotoxic agents
Central Nervous System Malignancies
Biological therapy
Presenter Kathryn Field
Authors K. Field1, M.A. Rosenthal1, H. Wheeler2, L. Cher3, E. Hovey4, A.K. Nowak5, C. Brown6, A. Livingstone6, K. Sawkins6, R.J. Simes6
  • 1Department Of Medical Oncology, Royal Melbourne Hospital, 3050 - Parkville/AU
  • 2Medical Oncology, Royal North Shore Hospital, Sydney/AU
  • 3Neuro-oncology, Austin Hospital, Melbourne/AU
  • 4Medical Oncology, Prince of Wales Hospital, Sydney/AU
  • 5Medical Oncology, Sir Charles Gairdner Hospital, Perth/AU
  • 6Clinical Trials, NH&MRC Clinical Trials Centre, University of Sydney, Sydney/AU



Glioblastoma (GBM) is the most aggressive malignant glial tumor. There is no accepted standard management after disease progression. Much remains unknown about the optimal use of bevacizumab (bev), including the role of continuing beyond progression, and patterns of radiological progression during and after use. In addition, the new Response Assessment in Neuro-Oncology (RANO) criteria have yet to be validated prospectively.


This study is a multi-centre, stratified randomised phase II trial. Patients have recurrent GBM after radiotherapy and temozolomide, have had no other chemotherapy for GBM, and have ECOG performance status 0–2. At least three months have elapsed since radiotherapy. In Part 1, patients are randomised 1:1 to intravenous bev 10 mg/kg 2-weekly and carboplatin AUC5 4-weekly or bev monotherapy. On progression, eligible patients are randomised to continue or cease bev (Part 2). The primary objective is to determine the effect of bev plus carboplatin versus bev alone on progression-free survival according to modified RANO criteria. Secondary endpoints are response rates, cognitive function, quality of life, steroid dose, toxicity, and overall survival. CogState, a validated neurocognitive testing system, is being used prospectively for the first time in this population and compared with mini-mental state examination. Exploratory endpoints include biomarker analyses, comparison of modified RANO and modified MacDonald criteria, predictive value of early MRI after 2 doses of bev, steroid dosing, and location and type of radiological progression during and after therapy.


Enrolment commenced in Nov 2010, completing Part 1 accrual in Mar 2012 with 122 patients randomised from 17 Australian sites. Randomisation to Part 2 continues. Feasibility and safety data will be presented; efficacy outcome results are not expected until 2013.


The study results will significantly improve knowledge regarding the use of bevacizumab in the setting of recurrent GBM, as well as providing for the first time a prospective analysis of CogState neurocognitive testing for patients with brain tumours.


M.A. Rosenthal: Roche Advisory Board member.

H. Wheeler: Roche Advisory Board member.

E. Hovey: Roche Advisory Board member.

A.K. Nowak: Roche Advisory Board member and has received research funding through Roche Australia.

R.J. Simes: Roche Advisory Board Member.

All other authors have declared no conflicts of interest.