Skip to content

Emulating randomized controlled trials with hybrid control arms in oncology: a case study

Published

January 2023

Citation

Sengupta, S., Ntambwe, I., Tan, K., Liang, Q., Paulucci, D., Castellanos, E., Fiore, J., Lane, S., Micsinai Balan, M., Viraswami-Apanna, K., Sethuraman, V., Samant, M. and Tiwari, R. Emulating Randomized Controlled Trials with Hybrid Control Arms in Oncology: A Case Study. Clin Pharmacol Ther. (2023) https://doi.org/10.1002/cpt.2841

Our summary

Randomized controlled trials (RCTs) are considered the gold standard to measure treatment effectiveness because randomization helps to reduce bias by balancing measured and unmeasured confounders across treatment groups. However, RCTs can be expensive, time-consuming, and challenging to conduct in some populations.

Real-world data (RWD) can be used to create a comparison group in situations where an RCT is not practical or ethical; however, there may be concerns with residual unmeasured confounding with fully external control arms. Hybrid control arms, which combine prospectively-planned clinical trials with real-world patients, offer a potential way to use RWD to increase power and save time while maintaining the benefits of randomization.

In this proof-of-concept study, researchers from Bristol Myers Squibb and Flatiron Health combined a hybrid control arm design with real data from previous phase 3 RCTs and RWD in oncology to emulate the treatment effect and estimate timeline savings compared to traditional RCT methods.

Why this matters

Incorporating RWD into clinical trials, such as through a hybrid control arm, offers several benefits in the context of clinical research - it can potentially lead to increased cost savings and faster decision-making, making treatments more widely available to patients. Hybrid study designs can also expand the breadth of patients included, providing more diversity in the trial population. Additionally, it increases patients' chances of being randomized to the investigational arm, sparing them from receiving less effective or less desirable therapy in the control arm.

Read the research

Share