Upstream Versus Downstream Complement Blockade: Potential Mechanistic Advantages of aC1s Targeting in Myasthenia Gravis
Shahar Shelly1, Marianna Lalla2, Yang Zhao2, Linda Rehaume2, Jennifer Cross2, Tuan Vu3
1Rambam Medical Center, 2Dianthus Therapeutics, 3University of South Florida
Objective:
To determine whether active C1s (aC1s) inhibition by claseprubart (DNTH103) suppresses generation of C3a, C3b and MAC in comparison with C5 inhibitor ravulizumab.
Background:

In generalized myasthenia gravis (gMG), pathogenic antibodies activate the classical complement pathway, culminating in membrane attack complex (MAC)-mediated injury at the neuromuscular junction (NMJ). Levels of complement components (C3) have been associated with disease severity in animal models. C5 inhibitors block MAC formation and improve clinical outcomes, without reducing the upstream inflammatory fragments (C3a, C3b), which potentially drive NMJ injury and amplify immune responses in addition to MAC formation.

Design/Methods:

MAC formation was quantified using the Wieslab® Complement Classical Pathway assay in 1% normal human serum (NHS). C3a in supernatants was measured by ELISA and C3b deposition on sensitized human red blood cells in 5% NHS was assessed by flow cytometry. For each assay, we tested dilution series of claseprubart, ravulizumab, and an isotype control.

Results:

In head-to-head assays, claseprubart and ravulizumab produced comparable inhibition of MAC formation, confirming similar downstream blockade. Only upstream aC1s inhibition with claseprubart resulted in near-complete reduction of C3a and C3b levels, while ravulizumab did not. Findings were reproduced across three independent experiments, supporting a mechanistic upstream (aC1s)-downstream (C5) inhibition distinction.

Conclusions:
Upstream aC1s inhibition with claseprubart prevented MAC formation and reduced C3a and C3b levels, demonstrating broader inflammatory control beyond C5 blockade. Thus upstream inhibition may provide both pathological and immunological advantages.
Generative AI Usage
No, did not use generative AI in the drafting or editing in this abstract.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.