# Emulating real-world GLP-1 efficacy in type 2 diabetes through causal learning and virtual patients

## TL;DR

This paper introduces an AI-based method to simulate clinical trials using real-world data, showing it can accurately predict treatment effects for type 2 diabetes.

## Contribution

A novel AI-driven approach that combines causal learning and generative models to emulate clinical trials and extrapolate results to broader populations.

## Key findings

- The virtual trial accurately predicted HbA1c reduction for GLP-1 compared to insulin and placebo, matching the LEAD-5 trial results.
- The model demonstrated significant differences in treatment outcomes using a difference-in-differences approach.
- The AI method is generalizable and can explore counterfactual scenarios for broader clinical decision-making.

## Abstract

Randomized controlled trials (RCTs) remain the benchmark for assessing treatment effects but are limited to phenotypically narrow populations by design. We introduce a novel generative artificial intelligence (AI) driven emulation method that infers effect size through virtual clinical trials, which can emulate the RCT process and potentially extrapolate into wider populations. We validate the virtual trials by comparing the predicted impact of glucagon-like peptide-1 (GLP-1) agonists on HbA1c in type-2 diabetes (T2DM) with its true efficacy established in the LEAD-5 trial. Our emulation model learns treatment effects from real-world evidence data by a combined generative AI and causal learning approach. Training data comprised pre- and post-treatment outcomes for 5,476 people with T2DM. We considered three treatment arms: GLP-1 (Liraglutide), basal insulin (glargine), and placebo. After training, virtual trials were conducted by sampling 232 virtual patients per arm (according to the LEAD-5 inclusion criteria) and predicting post-treatment outcomes. We used difference-in-differences (DiD) for pairwise comparisons between arms. Our goal was to emulate LEAD-5 by demonstrating a significant DiD in post-treatment HbA1c reduction for GLP-1 compared to basal insulin and placebo. We found significant differences in HbA1c reduction for GLP-1 vs basal insulin (-1.21 mmol/mol (-0.11%); p < 0.001) and GLP-1 vs placebo (-2.58 mmol/mol (-0.24%); p < 0.001) in our virtual populations, consistent with LEAD-5 (Liraglutide vs glargine: -2.62mmol/mol (-0.24%); p = 0.0015, Liraglutide vs placebo: -11.91 mmol/mol (-1.09%); p < 0.0001). The causal AI-powered clinical trials can emulate LEAD-5 in important measurements for T2DM. Our algorithm is specialty agnostic and can explore counterfactual questions, making it suitable for further study in the generalizability of RCT results in real-world populations to support clinical decision-making and policy recommendations.

Clinical trials are the gold standard for evaluating how well treatments work, but they have limitations. These trials typically include a very specific group of patients, meaning their findings may not fully apply to the broader population. They are also expensive and time-consuming. To address these challenges, we developed an artificial intelligence-based approach that can simulate clinical trials using real-world data. Our model generates virtual patient populations that reflect real-world diversity, allowing us to estimate how treatments like GLP-1 receptor agonists impact blood sugar levels in people with Type 2 diabetes. By using a combination of causal discovery and machine learning, our approach mimics the process of a traditional clinical trial while adjusting for key factors that influence treatment outcomes. We tested our method by replicating the results of a well-known diabetes trial (LEAD-5) and found that our virtual trial produced similar findings. This suggests that the causal AI-powered trial emulation could help extend clinical trial results to a wider range of patients, supporting more personalized treatment decisions and improving healthcare policy. Our approach could be applied to other diseases, providing a powerful tool for medical research beyond diabetes.