Researchers from Canada and Italy developed a new model to estimate the effect of treatment on the progression of disability in multiple sclerosis (MS). The model could help identify patients who are most likely to be responsive to treatment so that they can be preferentially included in clinical trials testing potential new treatments.

The team of researchers, led by Douglas Lorne Arnold, MD, FRCP, from Montreal Neurological Institute, Department of Neurology and Neurosurgery, McGill University in Quebec, Canada, used a multiheaded multilayer perceptron to estimate the conditional average treatment effect. The researchers gathered data from 6 randomized clinical trials that included 3830 patients. They first pretrained their model on a subset of patients with relapsing-remitting MS. Then, they fine-tuned the model on a subset of patients with primary progressive MS.

The study is published in Nature Communications.


Continue Reading

Read more about the different types of MS

When they used their model on a separate set of 297 patients with primary progressive MS who were either treated with anti-CD20 antibodies or given a placebo, they saw that the average treatment effect was larger for patients predicted to be most responsive using the model than for the entire group.

The researchers were also able to identify responders using their model in another group of 318 patients with primary progressive MS who were treated with laquinimod. 

“ . . . we show that using this model for predictive enrichment results in important increases in power,” the researchers wrote.

Multiple sclerosis is a progressive neurodegenerative disease characterized by an autoimmune attack on the myelin sheath, leading to disability among other symptoms. Available treatments are not able to halt the progression of disability, and there are currently no reliable biomarkers that can predict the effect of novel treatments on the progression of disability in patients with MS. 

Reference

Falet JPR, Durso-Finley J, Nichyporuk B, et al. Estimating individual treatment effect on disability progression in multiple sclerosis using deep learning. Nat Commun. 2022;13(1):5645. doi:10.1038/s41467-022-33269-x