Myasthenia gravis: From single cell signatures to cancer diagnosis

Abstract

Myasthenia gravis is a rare but archetypic autoimmune disease that is characterized by the autoantibody-mediated disruption of the neuromuscular junction leading to a skeletal muscle weakness. Immunomodulatory treatment options for Myasthenia gravis patients are largely unspecific, include suppression of the entire immune compartment and are often accompanied by severe side effects.

In order to identify novel biomarkers for more targeted and effective therapeutic approaches, we combined high-dimensional mass and flow cytometry with supervised and unsupervised machine-learning algorithms. Analysis of the peripheral immune compartment of Myasthenia gravis patients and healthy controls revealed a cellular immune signature consisting of inflammatory memory T helper cells with a defined cytokine profile. The abundance of the identified leukocytes in the blood strongly correlated with the patients clinical disease activity, far better than auto-Ab titers. Moreover, we were able to locate T cells with the defined signature enriched in the inflamed thymus of Myasthenia gravis patients – the key organ for the induction and maintenance of the autoimmune disease. Lastly, using an unbiased pattern recognition approach, we identified lymphomas in a subset of Myasthenia gravis patients, further highlighting the potential of the applied analysis tools.