OBJECTIVES: Randomised controlled trials (RCTs) with placebo controls often face challenges in demonstrating superior benefits over standard of care (SoC). This research aims to explore the potential benefits and risks of using AI-driven synthetic control arms (SCAs) as an alternative to placebo comparison data in clinical trials.

METHODS: Four trials evaluating the validity of “digital twins” (NCT04203823, NCT04849923, NCT05313594, NCT05181449) as SCAs were identified and reviewed. The application of "AI-driven advanced analytics" in using historical clinical trial data and real-world evidence (RWE) to create SCAs for the regulatory assessments of avelumab, blinatumomab, and the indication expansion of alectinib, was also investigated.

RESULTS: Two digital twin validation trials were ongoing, and two were completed. Anatomically-based modelling approaches were used to predict personalised treatment outcomes. Results from one completed trial showed statistically significant achievement of the primary endpoint in Type 1 Diabetes.

Blinatumomab, avelumab, and alectinib target leukemia, Merkel cell carcinoma, and ALK-positive non-small cell lung cancer, respectively. Large datasets of patient data from previous studies with existing treatments or standard of care and patient records were analysed using AI-produced algorithms to create historical control groups. Comparative effectiveness of blinatumomab, avelumab, and alectinib against these control groups, along with additional evidence, supported the establishment of clinical benefit and subsequent regulatory approvals.

CONCLUSIONS: The application of digital twins and AI-driven advanced analytics for SCAs offer potential benefits for comparing new drug efficacy against SoC. They could also address ethical and recruitment challenges in placebo-controlled trials by enabling robust sample sizes in rare and severe disease studies. However, digital twins are in validation stages with preliminary results, and the use of RWE and historical data for SCAs has been accepted as supplementary evidence alongside placebo RCTs. Further enhancing data quality and accuracy and establishing regulatory guidelines will be crucial to enhance confidence in these approaches in the future.