Hello everybody, my name is Saraf Salouki. I am an MD-PhD student from the Erasmus Medical Centre in Rotterdam. I'm very happy to talk to you today about dorsal root ganglion or DRG stimulation as a novel target for neuromodulation in spinal cord injury. I have no relevant conflicts of interest to declare. In this talk I would like to first of all explain why it would be relevant to look for a new target for neuromodulation in spinal cord injury in the first place. Then I'd like to explain to you the first steps that we took in order to prove the principle of this new target. And finally, I want to zoom in on what we think are relevant and important future steps to have this new target reach actual clinical maturity. So if we talk about neuromodulation in spinal cord injury research, we have to mention spinal cord stimulation or SCS, which involves the placement of electrodes over the epidural aspect of the spinal cord. And what this technique does is actually, through the recruitment of afferent pathways, activate relevant ventral motor neuron pools and their respective muscles. This is a video by the group of Courten in Switzerland who are actually very far in the application of spinal cord stimulation. It demonstrates how carefully these epidural electrodes are placed in an intraoperative setting. The placement of these electrodes, amongst many other things, dictates how selective the actual muscle recruitment is. The importance of this selectivity becomes clear when you look at the application after the surgery, where in combination with rehabilitation, a spatial temporally accurate locomotion pattern is being recreated using these electrodes. So the selectivity is necessary to actually target the right muscles at the right time without targeting other muscles that are unwanted. So interestingly, this selectivity has to be achieved in an anatomically very challenging area. The dorsal aspect of the spinal cord receives inputs from spinal roots from many different spinal cord levels, entering in very close proximity to each other. SCS started out as a chronic pain treatment, actually. And not too long ago, a new chronic pain treatment arose called dorsal ganglion or DRG stimulation. And as you can see, DRG stimulation has a target downstream from the dorsal aspect of the spinal cord. And this target is anatomically well separated. Each spinal cord level has their DRG in their own neuroforamen. So the idea arose that this inter-DRG selectivity that is seen in chronic pain treatment could also be potentially beneficial for locomotion or spinal cord injury research. In addition to this inter-DRG selectivity that could be potentially beneficial, DRG stimulation versus spinal cord stim in chronic pain has also shown to be able to reach what we call subdermal selectivity, so actually intra-DRG selectivity. What is more, the placement of the DRG leads involves a minimally invasive procedure in contrast to spinal cord stimulation. And finally, due to the placement of the DRG lead in the neuroforamen, you can first of all reach a more physically stable situation as compared to the epidural space used for spinal cord stimulation. And secondly, due to the fact that there is not much cerebral spinal fluid surrounding the DRG, you can actually reach good tissue lead contact. So to summarize, chronic pain treatment tells us that there are several unique benefits of DRG stimulation, which could also be potentially beneficial for locomotion or spinal cord injury research. So our group set out to perform a two-stage proof-of-principle study in order to look at this new target. In the first stage, pilot 1, we included patients with chronic pain already implanted with a DRG device, and we set out to develop a protocol to determine under which parameters we could elicit these motor responses. I will not be discussing this exact protocol in this talk due to time constraints. You are very welcome to email us in case you have questions about this. What I will be discussing is pilot 2, where we studied spinal cord injury patients applying this actual protocol and looking at whether DRG stimulation could elicit selective and clinically relevant lower extremity motor response. In this study, we included a total of 5 patients with motor-complete chronic spinal cord injury who were not implanted with a Baclofen pump. And these patients received bilateral DRG leads on the L4 level DRG. All of these were temporary, meaning that they were actually externalized. And the reason why we chose the L4 level specifically is because we were interested in invoking a motor response that would be already clinically relevant, and we in this case aimed for a knee extension movement, which would also allow for, for example, standing. These patients were implanted on day 1, and they were measured on day 1 as well as day 5, which was also the day of lead removal. You can see that we had a total of 4 outcome measurements that I will be discussing in the rest of this talk. What is important to mention is that we also had a sub-threshold stimulation period between day 1 and day 5, where patients were stimulated at a threshold below the motor response to see if we could have any neuromodulatory effects during the days that they were at home. First, a quick overview of the patients that we actually included. They were all male and relatively young, as you can see here. And important to note, all patients had a baseline MRC score of 0, indicating that indeed they were motor-complete spinal cord injury patients. Before diving into each of the outcome measures, here is an overview of patient 1 on day 1 and day 5. On day 1, we evoked isokinetic motor responses, a low-frequency dynamic motor response, as you can see here. We also aimed to evoke isotonic motor responses. These are strong, long-lasting, continuous muscle contractions, here being scored using MRC scoring. These isotonic motor responses could also be evoked bilaterally, as we did here on day 5. And as you can see, it gives a very clear knee extension. This knee extension, bilaterally, was then also used on day 5 in a, what we called, assisted weight-bearing ability test. The patient here was placed in a passive patient-diff system. And with the help of parallel bars and his upper extremity strength, you will see in a second that he is able to take weight on both legs. However, you will also see this is not completely stable, because the patient is, of course, lacking postural stability as well as hip stability. So, if we dive into each of the outcome measures, we start with the EMG traces, of which we calculated the average muscle power using the RMS, or root-mean-square value, both for isokinetic and isotonic responses, as you can see here. So, in these box plots, we displayed group-level data on both day 1 and day 5, with, on the y-axis, the average power in microvolts, and on the x-axis, each of the muscles on day 1 and day 5. We also calculated significance in two ways. The first thing we were interested in is which of the muscles were significantly activated as compared to their contralateral counterparts. So, as you can see here, it is mostly the upper leg muscles involved, also in a strong knee extension movement, that were significantly activated ipsilaterally to the side of stimulation, as compared to their contralateral counterpart. This in contrast to, actually, the lower leg muscles. Additionally, we wanted to look at the reproducibility of these responses on day 1 versus day 5. And on both isokinetic and isotonic stimulation, we see that for none of the muscles there was a significant difference in their activation on day 1 versus day 5, which means that, actually, the responses that were produced were stable. We also looked at the MRC scores during isotonic stimulation. As you can see here, both on day 1 and day 5, we were able to evoke isotonic motor responses with an MRC score of 4 or higher, indicating that these muscle responses leading to knee extension could be potentially weight-bearing. So, we also looked at to what extent patients were able to come to assisted weight-bearing, and this was the case for mostly patient 1 and 2, and to some extent patient 3. And this correlated with to which extent these patients also had arm strength and remaining core stability, as you can expect based on what you saw in the video of patient 1. We also looked at the patient diary, which was especially relevant for those days in between hospital measurements, where patients were at home and stimulated at a submotor threshold level. We saw two important responses. Three patients mentioned something that came close to an increase in core stability, with one patient describing that as feeling more tension in their underbelly and hips, and being able to sit better in their wheelchair. Additionally, two patients mentioned a decrease in severity and frequency of spasticity-related complaints, and one patient described that as feeling that the spasticity they would experience when they changed position was actually reduced during stimulation. So, if we summarize the outcomes of this study as follows, we can conclude the following things. Selectivity-wise, we can indeed see that unilateral L4-DRG stimulation gives an L4-specific EMG response, which is also selective functionally, meaning that it gives a very clear knee extension movement. This was also reproducible and stable over the period of 5 days. Additionally, we have seen some clinically relevant responses. On the one hand, we're able to stimulate both isokinetic and isotonic responses, which have their own relevance in the clinic. We've also seen that the isotonic responses are indeed potentially weight-bearing, although we need more developments for that. And finally, we have seen some hints for other beneficial side effects due to DRG stimulation. So, knowing these results, how far are we actually from clinical application? Well, compared to SES, we're actually still quite far, especially because of the fact that hardware out there does not facilitate the multichannel stimulation that we would need to recruit all the relevant DRGs involved in locomotion at once. Additionally, in this study, we only looked at stimulator-evoked motor responses. And of course, the ultimate goal would be to consider the electrical stimulation in combination with actual rehabilitation protocols, to see if the selectivity that we can reach with DRG also reflects back in an actual functional and clinical improvement during these rehabilitation protocols. However, we also think that there are more short-term benefits to what we have shown here today, one of them being the possibility of synergistic DRG and spinal cord stimulation. Applying DRG stimulation, for example, to those areas is particularly difficult for SES to reach in a selective manner, one of them being the sacral part of the spinal cord. We could place DRG leads there, and upstream to that, continue placing the original SES electrodes to use both techniques in a combined manner. For now, I'd like to thank the multidisciplinary DRG team, consisting of experts in engineering, neuroscience, neurosurgery. I would like to thank you for your attention, and would like to invite you also to ask any questions you have through the email address displayed here. And finally, these are the references, in case you are interested. Thank you again.

dorsal root ganglion stimulation

neuromodulation

spinal cord injury

chronic pain treatment

locomotion enhancement