The definition of platinum resistance was made more specific: progression within < 6 months from completion of a minimum of 4 platinum therapy cycles, with the date calculated from the last administered dose of platinum therapy. The definition of prior therapies was clarified to include all previous anti-cancer therapy, including those received in the front-line or recurrent settings. It was clarified that it was not a requirement in RECIST for the same investigator to evaluate the participant at each assessment. The statistical analysis for the primary endpoint was updated from a one- to a two-sided log-rank test. The numbers of participants randomized to chemotherapy cohorts was amended because of statistical changes to include 120 participants per chemotherapy cohort. The timing of QoL assessments was amended to be more suited to the scheduling of cycle visits, the 3 worst symptom questionnaire was collected at baseline only, and the use of the 3 worst symptom questionnaire methodology was described in more detail. The protocol was updated so that all Grade 2 adverse events were collected. It was clarified how a participant who had been previously enrolled in a blinded study with an anti-angiogenic was to be stratified. The frequency of CA-125 assessments was corrected to be performed every cycle, not at every visit. It was clarified that the “optional post-study phase” for participants randomized to the CT arm was for the CT arm only and that bevacizumab was to be given as part of the study to those participants who opted to receive crossover bevacizumab monotherapy. Additional safety guidance was provided for the management of bevacizumab in the event of CNS bleeding, proteinuria management, and hypersensitivity with paclitaxel. The definition of residual disease was amended based on the presence or absence of macroscopic disease. Definitions of progression for participants with measurable and non-measurable disease at randomization were further detailed.

Clarification regarding the exclusion criteria for platinum refractory disease, peripheral neuropathy, and previous malignancies. Addition of left ventricular ejection fraction (LVEF) assessments every fourth cycle for participants receiving pegylated liposomal doxorubicin (PLD). Additional requirement to capture certain concomitant medication in the electronic Case Report Form (eCRF), particularly supportive medication prescribed for the treatment of cancer-related symptoms or potential side effects of chemotherapy. Clarification that only serious adverse events caused by protocol-mandated interventions needed to be collected prior to initiation of study medication and that all serious adverse events needed to be collected before, during, and after study drug dosing. Guidance on dose modification to reflect the bevacizumab safety profile. Clarification to ensure that only those participants who experienced disease progression on chemotherapy alone were able to subsequently receive bevacizumab on the bevacizumab crossover option.

Allow for a potential retrospective scan collection and a review of scans by an independent review committee (IRC). Clarify that the duration of survival follow-up should continue for a minimum of 12 months after end of treatment for all participants.

The amendment defined that the study would be closed as soon as the protocol amendment was approved by regulatory authorities and Ethics Committees. The amendment clarified that participants who were still receiving investigational study medication (bevacizumab) would end their AURELIA study participation. If the Avastin Long Term Extension study (AvaLTE, MO25757) was approved in the participant’s country, the participant would be offered participation in this study. Alternatively the participant would be offered continued bevacizumab treatment with commercial drug until disease progression, unacceptable toxicity or participant request for discontinuation as initially planned by the protocol.