Our study represents a milestone that establishes, for the first time, a correlation between multimodal electrophysiology data (neuronal discharge characteristics in MER, stimulation dependent frequency band power changes in scalp EEG) and intrathalamic white matter tract volumes at corresponding sites, in general epilepsy patients. We observed that stimulating CM target sites with lower neural activity increases delta band power. These targets, located adjacent to the CM in the internal medullary lamina, provide mechanistic insights into the effects of CM neuromodulation, potentially elucidating its underlying mechanisms of action. This novel approach may be a crucial step towards establishing comprehensive guidelines for intraoperative electrophysiological guided implantation of thalamic stimulation electrodes in patients with drug-resistant generalized epilepsy.
Here we demonstrate that even small changes in depth along a tract can show significant differences in neural activity and stimulation of these various targets can produce differences in delta band power changes in the scalp EEG. We then demonstrate that axonal stimulation corresponding to the IML, rather than cell body stimulation, are associated with greater cortical electrophysiological responses in the delta frequency range. Our method provides real-time feedback of intraoperative neurophysiology can assist the surgeon and clinical team in making immediate adjustments and improving the accuracy of depth electrode placement which would be critical for improving seizure outcomes13. Furthermore, targeting of white matter tracts diverges from conventional and well-established MER guided techniques for DBS implantation. For example, the placement of DBS electrodes in patients with Parkinson’s disease is targeted to regions of the STN with the highest cell density by unitary recordings 23. Here, the opposite may be true for CM stimulation for generalized epilepsy, as we find that targets with low cell density, likely the IML white matter bundle surrounding the nucleus, could provide more beneficial effects.
The IML is an important thalamic structure that has yet to be fully studied in reducing seizures. It is a thin layer of white matter that runs through the center of the thalamus, dividing it into medial and lateral parts. It consists of myelinated nerve fibers that surround the CM and project both to and from various parts of the cerebral cortex. The IML plays an integral role in the organization and connectivity of the thalamus. It serves as a crucial relay station for sensory and motor signals in the brain. As part of the reticular formation, the IML also is responsible for regulating levels of cortical excitability, influencing, and changes in cortical activity, as observed in sleep and wakefulness cycle. It also plays a likely role in epileptic mechanisms related to loss of consciousness and alertness. Loss of consciousness and alertness are key factors associated with morbidity and deterioration of quality of life in patients with generalized epilepsy.
Further investigation is warranted to replicate our results of a widely distributed increase in delta power following CM stimulation. In this study, the sole consistent EEG frequency modulation that increased in power with CM stimulation was observed in the delta-band activity. Studies have demonstrated that cortical delta activity is thalamus-driven, occurring when thalamocortical neurons transition to a burst mode of firing, effectively isolating cortical neurons from sensory input27,28. Thalamic bursting, induced by rebound excitation when thalamocortical neurons are rapidly released from inhibition, occurs at delta frequencies. This process may toggle cortical pyramidal cells between up and down states at the same frequency while obstructing ascending sensory input29,30. Noteworthy is the correlation between the deactivation of the thalamus, measured by regional cerebral blood flow, and the emergence of delta oscillations31. Taking this into consideration, we assert that the increase in delta power plays a pivotal role in locating an optimal target or “sweet spot” within the CM for thalamic neuromodulation. This proposal suggests that the ability for fine adjustments of the final implant location based on electrophysiological data, such as modulated delta power in the EEG, can potentially offer greater disruption of the epileptogenic network. Our study may indicate that the optimal target in CM neuromodulation for epilepsy in the adjacent IML and not the CM per se. Future studies are required to see if this can enhance the clinical efficacy of neuromodulation in reducing seizures.
Our study has limitations. While it is based on intraoperative findings and acute changes under sedation, the robustness and consistency of our results are noteworthy. The effectiveness of CM stimulation guided by the described method in long-term seizure control and in a larger cohort of patients warrants further investigation. Additionally, although our results were acquired from a relatively small cohort of patients, the use of several recording and stimulating sites (36 in total) demonstrates high consistency and robustness, enhancing the reliability of our findings.
In conclusion, the results of different multi-unit recordings and electrical stimulation locations within the CM nucleus and adjacent areas demonstrated that the targets with lower cell density, potentially indicating axon predominant regions, were associated with widespread cortical delta band frequency changes. We observed that the thalamic area corresponding to the lower cell density regions is the IML. These findings underscore the imperative for advancing intraoperative electrophysiology to identify and guide implantation in new thalamic targets, such as the IML, thereby enhancing the efficacy of neuromodulation in treating generalized epilepsy.