To comply with updated regulations, serious adverse events (SAEs) and pregnancies should be reported within 24 hours. Instructions were provided on how to record when a previously recorded non-serious adverse event becomes an SAE. Clarification was provided regarding assessments to be performed at the study drug discontinuation visit for subjects who have been unblinded for potential crossover and are subsequently treated with bevacizumab in the study. To allow sufficient time to address certain adverse events and to monitor the effect of medical management, the maximum time allowed for delay of study drug administration was extended from 7 days to 21 days. The window for pregnancy testing at screening was changed from 48 hours to 7 days to ensure that the test is performed prior to study drug administration and to align with the protocol’s schedule of assessments. The instructions for study drug delay/discontinuation due to proteinuria were corrected to align with the rest of the protocol as well as the bevacizumab Summary of Product Characteristics. The specific instructions for study drug discontinuation due to fistula were clarified.

The enrollment into Treatment Arm C (onartuzumab + placebo) was permanently discontinued. A temporary suspension of enrollment into Arm C was announced via a letter to investigators dated 21 September 2012, pending the investigation of a potential safety concern of cerebral oedema in four out of five subjects in Arm C. During the suspension, randomisation to Treatment Arm A (bevacizumab + placebo) and Arm B (bevacizumab + onartuzumab) continued. On 17 October 2012, the Internal Monitoring Committee (IMC) and Scientific Oversight Committee (SOC) concluded that the cerebral oedema events reported in these subjects were findings consistent with the documented progression of their underlying glioblastoma and that there was no clear attribution to study drug. It was then determined that an insufficient number of subjects would be enrolled in Arm C compared to the other two arms by the targeted enrollment completion of March 2013 while maintaining a 1:1:1 randomisation.. Based on this assessment, the Sponsor decided to permanently discontinue enrollment into Arm C.

Amendments were made to reduce the protocol-specified assessments for subjects who were either receiving study treatment or were in the survival follow-up period following the primary efficacy analysis and unblinding of treatment assignments. The Sponsor recommended that subjects continuing to receive study treatment discontinue onartuzumab. The decision to continue or discontinue treatment with onartuzumab was at the discretion of the investigator in consultation with the subject. Bevacizumab treatment could continue until there was evidence of progressive disease, until treatment-limiting toxicity developed, until the treating physician considered the subject to no longer be achieving benefit, or until the patient decided to withdraw, whichever occurred first. This protocol amendment reduced the number and type of assessments required for evaluation of study treatment but retained the requirements for submission of serious adverse event data. Important safety information regarding venous thrombosis and embolism was updated. The signs and symptoms of venous thrombosis and embolism were to be carefully monitored. The investigator was responsible for instructing subjects to seek medical care if they developed symptoms, such as shortness of breath, chest pain, or arm or leg swelling. Thromboprophylaxis had to be considered after careful assessment of an individual subject’s underlying risk factors. Any thromboprophylaxis had to follow the current institutional guidelines and best practice.