Unlearn, an artificial intelligence company, has partnered with APST Research to enhance clinical trials for amyotrophic lateral sclerosis (ALS) by creating digital twins of patients. The collaboration leverages APST's database of over 8,000 ALS patients to train Unlearn's Digital Twin Generator (DTG), an AI platform designed to predict individual patient outcomes under control treatments.
The DTG aims to improve statistical power in clinical trials, identify subgroups with varying treatment responses, and add simulated comparator groups. This approach could potentially reduce the number of patients needed in control groups and accelerate research timelines for ALS, a progressive neurological disorder affecting motor neurons.
Leveraging Extensive ALS Patient Data
APST Research, collaborating with ALS centers in Europe, has compiled a dataset that includes standard clinical assessments such as the ALS Functional Rating Scale Revised (ALSFRS-R), lung function tests, and muscle strength measurements. The dataset also features patient self-assessments and biomarker analyses, including neurofilament light chain (NfL), a key indicator of nerve damage and disease progression.
According to Unlearn CEO Steve Herne, the partnership with APST provides access to the most robust ALS dataset available, which is crucial for improving the accuracy of digital twins. Thomas Meyer, MD, founder of APST, emphasized that this collaboration marks the first time APST's dataset has been licensed, creating a unique opportunity to positively impact ALS clinical trials.
Applications of Digital Twins in ALS Research
The integration of APST's data into Unlearn's DTG is expected to provide unprecedented insight into ALS progression. Elevated levels of NfL have been linked to the progression of ALS and are now widely used as a surrogate marker of drug efficacy in early-stage trials. The detailed data on ALS progression rates, real-world healthcare data on symptomatic drug treatments, assistive technology devices, nutritional support, and ventilation therapies will further enhance the DTG's performance.
Unlearn and APST plan to publish studies on advancements in digital twin technology and its applications. Unlearn will also provide APST with digital twins of study participants to support further research and enhance the dataset's utility. The goal is to accelerate research projects and clinical trials, bringing the field closer to a holistic understanding of ALS and potential breakthroughs for people with the disease.
By utilizing digital twins, Unlearn's platform supports clinical trials from design and planning to execution and analysis. This enables trial sponsors to increase statistical power, reduce the number of patients randomized to control groups, identify subgroups with different responses to treatment, optimize eligibility criteria, add simulated comparator arms to single-arm trials, and estimate individual treatment effects.
