Estimating individual treatment effect on disability progression in multiple sclerosis using deep learning

• Published in: Nature communications Vol. 13; no. 1; p. 5645
• Main Authors: Falet, Jean-Pierre R., Durso-Finley, Joshua, Nichyporuk, Brennan, Schroeter, Julien, Bovis, Francesca, Sormani, Maria-Pia, Precup, Doina, Arbel, Tal, Arnold, Douglas Lorne
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 26.09.2022; Nature Publishing Group; Nature Portfolio
• Subjects: 692/308/2779; 692/617/375/1666; Antibodies; Biomarkers; CD20 antigen; Clinical trials; Deep learning; Drug development; Enrichment; Humanities and Social Sciences; multidisciplinary; Multilayer perceptrons; Multiple sclerosis; Patients; Science; Science (multidisciplinary); Test sets
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-022-33269-x

Disability progression in multiple sclerosis remains resistant to treatment. The absence of a suitable biomarker to allow for phase 2 clinical trials presents a high barrier for drug development. We propose to enable short proof-of-concept trials by increasing statistical power using a deep-learning predictive enrichment strategy. Specifically, a multi-headed multilayer perceptron is used to estimate the conditional average treatment effect (CATE) using baseline clinical and imaging features, and patients predicted to be most responsive are preferentially randomized into a trial. Leveraging data from six randomized clinical trials ( n  = 3,830), we first pre-trained the model on the subset of relapsing-remitting MS patients ( n  = 2,520), then fine-tuned it on a subset of primary progressive MS (PPMS) patients ( n  = 695). In a separate held-out test set of PPMS patients randomized to anti-CD20 antibodies or placebo ( n  = 297), the average treatment effect was larger for the 50% (HR, 0.492; 95% CI, 0.266-0.912; p  = 0.0218) and 30% (HR, 0.361; 95% CI, 0.165-0.79; p  = 0.008) predicted to be most responsive, compared to 0.743 (95% CI, 0.482-1.15; p  = 0.179) for the entire group. The same model could also identify responders to laquinimod in another held-out test set of PPMS patients ( n  = 318). Finally, we show that using this model for predictive enrichment results in important increases in power. There are limited predictive biomarkers for drug treatment responses in individuals with multiple sclerosis. Here using existing clinical trials data, the authors propose a deep-learning predictive enrichment strategy to identify which participants are most likely to respond to a treatment.