731P - REWENEC: REal-World Evidence in NEC provides a reliable external control arm to avoid randomization for second-line clinical trials - A post hoc proof of concept analysis with the randomized phase II BEVANEC study

Background

Neuroendocrine carcinomas (NECs) are rare aggressive cancers. After first-line platinum–etoposide, there is no standard second-line treatment. The phase II randomized clinical trial BEVANEC evaluated a randomized treatment arm (RTA) FOLFIRI + bevacizumab (Fb) and a randomized control arm (RCA) FOLFIRI (F) with a long period of enrolment (5 years) and demonstrated no benefit of Fb. We hypothesized that a single-arm trial of Fb compared to an external control arm (ECA) of F could have led to the same results.

Methods

We compared the patient’s characteristics and outcome in the RCA of BEVANEC versus a full or hybrid ECA generated with real world data of NEC patients from 2 retrospective French cohorts (CEPD and RBNEC) who received second-line F. We emulated a synthetic trial, REWENEC, incorporating an ECA of F and compared the results to that of BEVANEC. We assessed the potential impact of unmeasured confounders. An inverse probability weighting propensity score (IPW-PS) approach was used to balance measured confounders (MC) in the ECA.

Results

The ECA included 66 patients comparable to ETA and RCA. Measured confounders were more balanced between the RTA of BEVANEC and the ECA after IPW-PS than between the RTA and the RCA after randomization (Table). OS were similar in both ECA and RCA (Table). When substituting the RCA by the ECA and comparing it to the RTA, the synthetic REWENEC trial led to similar outcomes and conclusions as in the BEVANEC trial with no difference in median overall survival between Fb and F (Table). This suggest that randomization may not have been required in this setting.

Table: 731P

Conclusions

Externally controlled trials might be an appropriate compromise between RCT and single-arm trials in NECs. Maintaining a platform of existing good quality data from prospective and retrospective studies would help to design these types of trials and accelerate clinical research in NECs.

Clinical trial identification

NCT02820857.

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

J. Hadoux: Financial Interests, Personal, Advisory Board: Ipsen, Lilly, Pharma Mar; Financial Interests, Institutional, Invited Speaker: AAA, Pfizer; Financial Interests, Institutional, Advisory Board: EISAI, HRA pharma. A. Lievre: Financial Interests, Institutional, Funding: Bayer, Lilly, Novartis; Financial Interests, Advisory Board: AAA-Novartis, Astellas, BMS, Incyte, Pierre Fabre, Servier; Financial Interests, Invited Speaker: Amgen, Esteve, IPSEN, Leo Pharma, Mylan, Viatris. D. Tougeron: Financial Interests, Personal, Advisory Board: AstraZeneca, Sanofi, Amgen, MSD, Roche, Servier, Pierre Fabre, BMS, Bayer; Non-Financial Interests, Member of Board of Directors: Federation Francophone de Cancerologie Digestive. E. Baudin: Financial Interests, Personal, Other, Project lead and principal investigator: Ipsen; Financial Interests, Personal, Advisory Board: Novartis-AAA; Financial Interests, Institutional, Research Grant: Novartis, HRA; Non-Financial Interests, Principal Investigator: Enterome; Non-Financial Interests, Leadership Role, French network of endocrine and neuroendocrine tumors: Endocan Network. C. Lepage: Financial Interests, Personal, Advisory Board: AAA; Financial Interests, Personal, Invited Speaker: Amgen, Pierre Fabre, Ipsen. T. Walter: Financial Interests, Personal, Invited Speaker: Novartis-AAA, Ipsen; Non-Financial Interests, Institutional, Funding: Roche. All other authors have declared no conflicts of interest.