G-Ba Does not Adjust Evidence Requirements in Early Benefit Assessment in Cases of Pre-Defined, Efficacy-Based Cross-Over Decisions in Oncology Trials

In Germany, an early benefit assessment (EBA) by the Federal Joint Committee (G-BA) is compulsory for all new drugs. Pre-defined treatment switching, often called ‘cross-over’, is often seen in oncology clinical trials. Cross-over is usually implemented for ethical reasons, i.e. to ensure access to a beneficial treatment for all patients, but may confound data analysis by improving efficacy in the control arms. We aimed to analyse the impact of cross-over on evidence levels granted by the G-BA. Oncology medicines with completed EBAs by 01 Jan 2015 were analysed for i) presence of cross-over in pivotal trials; ii) efficacy results before and after cross-over and iii) evidence levels granted by the G-BA (proof, indication or hint). Cross-over was frequent in oncology, concerning 14 of 28 EBAs (50%). For 6 of the 14 medicines, cross-over could be considered ethically required as significant differences in overall survival (OS) were demonstrated prior to cross-over. For most medicines, data on OS and progression-free survival were reported after cross-over (10/14 and 8/14, respectively). Significant differences in OS post-cross-over could only be shown for 2 out of the 8 medicines for which no such differences were demonstrated before cross-over. An evidence level of proof was granted by the G-BA for 3 out of the 14 medicines, all of which were orphan drugs, but none were granted for medicines with ethically required cross-over. The G-BA regards evidence standards as only partially fulfilled in cases of ethically required cross-over in oncology. Highly efficacious drugs with ethically mandated cross-over are therefore systematically disadvantaged with regards to the achievable evidence category, indicating a bias against innovation. Medicines with a demonstration of superior efficacy and subsequent ethically justified cross-over deserve an evidence level of proof.