Background: AURELIA is a phase 3, randomized, open-label trial in patients (pts) with PT-resistant, recurrent OC (NCT00976911). Based on investigator (INV) assessment of radiographic data, CT + BV showed significant improvement vs CT alone in the primary end point of progression-free survival (PFS) (Pujade-Lauraine, JCO in press). Median PFS was 3.4 months for CT vs 6.8 months for CT + BV (stratified hazard ratio [HR], 0.38; 95% confidence interval [CI], 0.30–0.49; P<.001). As this was an open-label trial, an independent review committee (IRC) assessment of radiologic data was conducted retrospectively to confirm these results. Methods: Radiologic images were provided to the IRC for 92.2% of randomized pts; 82.5% of pts were IRC-evaluable. Data were reviewed in a blinded manner according to a prespecified charter following modified RECIST v.1.0. PFS analysis was based on the intent-to-treat population. Results: IRC results were consistent with those from the INV, demonstrating a statistically significant and clinically meaningful improvement in PFS with BV (Table). The concordance rate for progression status (yes, no) was similar across arms (CT, 69.9%; CT + BV, 68.2%), as was agreement on the date of progressive disease (PD; CT, 69.1%; CT + BV, 67.2%). The early discrepancy rates (INV-determined PD earlier than IRC-determined PD) were similar in the 2 arms (CT, 0.372; CT + BV, 0.364), whereas the late discrepancy rate was higher in the CT + BV arm (CT, 0.329, CT + BV, 0.463). The differential discordance of 0.134 suggests some degree of INV bias toward the CT + BV arm. Conclusions: The PFS analysis results from the IRC assessment were consistent with those from the primary INV-led PFS analysis, despite a trend for INV assessment of PD at a later date in the CT + BV arm in this trial. The results provided further evidence that INV-determined PFS based on RECIST is reliable and reproducible in OC clinical trials. Clinical trial information: NCT00976911.