From Velasco to NAUTILUS: an evolution of centromedian nucleus targeting in epilepsy

The field of neuromodulation for idiopathic generalized epilepsy is rapidly evolving. The recently completed NAUTILUS trial, an industry-sponsored multicenter, randomized, sham-controlled study evaluating centromedian nucleus (CM) stimulation, represents a landmark effort to rigorously assess this approach. Although trial results are not yet available, the need to understand and standardize the anatomical targeting of the CM remains critical. The precision of electrode placement fundamentally determines whether stimulation engages the intended thalamocortical circuits, shaping both efficacy and reproducibility across centers.

The authors present a comprehensive review of the evolution of CM targeting in epilepsy, tracing its development from early stereotactic approaches to current imaging-based strategies. The authors detail the surgical targeting workflow used across NAUTILUS trial sites, which incorporates advanced MRI sequences and atlas-informed refinements. This review also offers practical resources and tools enabling clinicians to implement the described methodology in clinical and research settings.

Modern CM targeting strategies reflect a transition from indirect, coordinate-based techniques to individualized, image-guided planning. These methods have been successfully implemented across centers participating in the NAUTILUS trial, enabling accurate electrode placement even in patients with complex anatomy. The workflow accommodates both direct visualization and atlas-based alternatives when imaging quality is suboptimal.

This review outlines the evolution of CM neuromodulation-from Velasco's early targeting approaches to modern image-guided surgery-and provides a reproducible framework for clinicians. As neuromodulation moves toward broader clinical adoption, anatomically precise targeting will be key to optimizing outcomes. The surgical approach recommended in the NAUTILUS trial offers an implantation framework intended to support future location-based outcome analysis.