Electrical interaction with the abnormal brain cavity wall to reduce cavity-related symptoms: proof-of-concept study in the rat

On each neurosurgery and neuroanatomy ward around the world lie patients with an abnormal brain cavity (aBC). Such an aBC develops because neurons are lost as a result of brain damage due to e.g. a stroke or traumatic brain injury. Symptoms associated with an aBC range from motor impairment such as paresis to psychiatric symptoms such as depression. Despite a wide range of available therapies, these symptoms often remain, even years after the damage first occurred. Over the years, new treatment strategies have thus been explored, especially for chronic stroke-related motor impairment. Electrical brain stimulation is one of those new treatment strategies. Current applications include direct electrical stimulation of the motor cortex and transcranial stimulation of the motor cortex. Both approaches have been shown to improve stroke-related motor impairment.

In this thesis, we propose a new strategy to treat aBC-related symptoms, namely via electrical stimulation of the aBC wall.

In Chapter 1, we demonstrated that stereotactic aspiration of the forelimb area of the motor cortex (fM) causes long-term motor impairment in rats. Furthermore, stereotactic aspiration also gave rise to a distinct cavity. We therefore concluded that this model can be used as a model for aBC associated with motor impairment.

In Chapter 2, we implanted a 4 x 4 electrode array against the aBC in this model for the first time. The electrode array was specifically designed for this project, and was ultrathin and flexible, making it ideal to implant against the curved aBC wall. We electrically stimulated the aBC wall with all electrical contacts simultaneously (i.e. non-selective stimulation), and evaluated the effect of various sets of stimulation parameters on motor performance. Electrical stimulation with a frequency of 100 Hz and a pulse width of 210 µs improved motor performance significantly in a subset of the most severely impaired rats. We could, however, not replicate these results, most likely due to technical issues. In a subsequent study, we changed our approach drastically and focused on an individualized approach in which the effect of non-selective and selective (i.e. with smaller groups of electrical contacts) electrical stimulation of the aBC wall on motor performance was investigated (Chapter 3). We evaluated reaching success and kinematic features of reaching and grasping, which were detected with an automated video-algorithm specifically developed for this project. We found that the effect of electrical stimulation was clearly specific for each subject. Furthermore, although electrical stimulation did not have a significant effect on reaching success, we did demonstrate that both non-selective and selective stimulation of the aBC wall influenced kinematic features of reaching and grasping significantly in 11 out of 13 rats. Interestingly, selective stimulation appeared to be more or at least as effective as non-selective stimulation, as either only selective stimulation elicited an effect, or the effect of both non-selective and selective stimulation was the same. Based on these results and observations from the literature, we hypothesize that the aBC wall consists of a mixture of hyper- and hypoactive regions, and that high-frequency electrical stimulation interacts with hyperactive regions thereby influencing the symptoms. Moreover, as we observed that selective stimulation was subject-specific, the neuronal activity in the aBC wall is most likely also subject-specific, and may depend on which neuronal circuitries are disrupted and where.

In Chapter 4, we investigated the neuronal activity in the aBC wall using local field potentials (LFPs). We demonstrated that the power of the theta and gamma frequency bands was significantly higher during activity compared to rest. These results demonstrate that (i) LFPs can be recorded from the aBC wall, and (ii) that LFP activity of the aBC wall is correlated to motor performance. We are currently also investigating LFP activity recorded from the aBC wall during reaching and grasping. We aim to correlate this LFP activity to kinematic features of reaching and grasping, and to compare the LFP activity between successful and failed reaching attempts. If we can predict when a failed attempt will occur, LFPs could be used to steer electrical stimulation of the aBC wall both in space and time (i.e. closed-loop stimulation). Closed-loop stimulation may be an important optimization strategy to further increase the effect size of the stimulation in future studies.

In Chapter 5, we investigated the feasibility and applicability of functional ultrasound imaging (fUSi) in this research context. Based on our preliminary findings, we conclude that fUSi can be used to investigate the neuronal activity in the aBC wall, and its response to electrical stimulation. This technique will thus be used in future experiments to better target electrical stimulation. Another part of optimization of our strategy is further improvement of the electrode design. In Chapter 6, we investigated whether a layer of extracellular matrix (ECM) proteins can be used as a temporary reinforcement of the electrode, and demonstrated that it is indeed suitable for this purpose. Temporarily reinforcing the electrode may facilitate implantation of ultrathin, flexible electrodes. Finally, optimal electrical stimulation may not only depend on where and when to stimulate, but also on how. Therefore, in Chapter 7, we evaluated the effect of motor cortex stimulation with complex pulse shapes on limb movement in anesthetized rats. We found that complex pulse shapes (e.g. biphasic pulse with interphase gap or a Gaussian pulse) are more efficient in evoking limb movement compared to the standard biphasic pulse.

We conclude that electrical stimulation of the aBC wall influences aBC-related symptoms significantly in a rat model for aBC associated with motor impairment, and thereby confirm our main hypothesis. Moreover, we showed that LFPs recorded from the aBC wall are correlated to motor performance, opening perspectives for the development of a closed-loop stimulation system. Finally, we presented potential strategies that may be used to optimize electrical stimulation of the aBC wall in the future.