Discover more from Unlearn
It’s NfL season: introducing our new Digital Twin Generator for amyotrophic lateral sclerosis
Daniele Bertolini, Machine Learning Scientist, and Jon Walsh, Head of Modeling
At Unlearn, our purpose is to advance AI to eliminate trial and error in medicine. This isn’t something we expect to accomplish overnight—but our pioneering work at the forefront of AI research today gives us confidence in the road ahead. With this mission in focus, we’re proud to announce the release of our latest Digital Twin Generator for Amyotrophic lateral sclerosis (ALS-DTG-2.2). This upgraded model builds upon its predecessor, boasting an expanded repertoire that now encompasses predictions for neurofilament light chain (NfL).
As you may have heard, NfL has generated a lot of excitement in the ALS community. NfL is a promising biomarker for identifying neurodegeneration in ALS, offering objective and quantifiable measurements in cerebrospinal fluid and blood as neurodegeneration occurs. This aids in assessing treatment effectiveness and disease severity, benefiting clinical trials and research. Additionally, NfL shows promise in early disease detection, potentially improving patient outcomes by identifying ALS before severe symptoms manifest.
As a biomarker for neurodegeneration, NfL is at the vanguard of clinical trial endpoints in ALS and an area of intense research. Case in point, Biogen’s drug tofersen was approved by EMA at the end of 2022 and granted accelerated approval by FDA earlier this year for a rare genetic form of ALS. A motivating factor in these approvals is the fact that the drug significantly reduced plasma NfL levels.
So, how does our new ALS-DTG-2.2 fit into cutting-edge NfL research? By generating digital twins of ALS trial participants that enable faster trials that maintain power with smaller control arms or highly powered trials without increasing sample sizes in both early stage and registration trials. This translates to enormous savings in both cost and time and, most importantly, allows more ALS trial participants to receive the experimental treatment.
In building this new model, we have incorporated data from ALS clinical studies that measure NfL and leveraged the same AI methodology that we’ve used previously in Alzheimer’s Disease, ALS, and Frontotemporal Dementia. The specification sheet for ALS-DTG-2.2 is available to download on our website, and it provides more information on model performance related to specific clinical trial endpoints.
At Unlearn, we build AI technology that leverages all types of health data to build comprehensive views of human health. The digital twins of patients we create today hold the key to unlocking the future of personalized medicine. We have created DTGs in ALS, Alzheimer’s Disease, and Frontotemporal Dementia, and we will continue to respond to advances in basic science and clinical trial designs, like the discovery and development of NfL as a biomarker, to bring new treatments to patients sooner.
ALS is ruthless, and we believe we must move fast to help defeat it. We build digital twins of trial participants that maximize the impact of their data by using it to run faster, more powerful clinical trials. Reach out to us today to learn about how to incorporate digital twins into your next clinical trial.