Hi, I'm Richard Schmidt. I'm one of the fifth-year residents at Thomas Jefferson University, and I'm presenting some data from our lab looking at a model for blood-brain barrier modulation using sphenopalatine ganglion stimulation. It's a very interesting project, a little bit of a background in the blood-brain barrier. Most people are pretty familiar with it, but just a quick recap. Essentially, very small lipophilic molecules can pass the blood-brain barrier and molecules with specific carrier proteins. Essentially, large lipophilic molecules and any hydrophilic molecules cannot get through. Essentially, it does a pretty good job of preventing toxins from entering the brain and foreign substances, but there are certain times when we would like certain things to get into the brain. Of course, therapies targeted brain tumors, chronic degenerative conditions, things of that nature. So right now, it's a major challenge to try and bypass this barrier. So a brief background on the sphenopalatine ganglion. A lot of the work in this was done in the mid-'90s and uncovered that it provides parasympathetic innervation to the anterior circulation. In lower mammals, particularly rats, it is a bilateral innervation. However, in higher mammals, this innervation is only on the ipsilateral side. Early attempts of doing electrical stimulation of this ganglion showed that it does exert a vasodilatory response in the anterior circulation territories, essentially showing in rats that when they stimulated the nerve, they were able to dilate the blood vessels and increase cerebral blood flow. And this was later done in monkeys, and then there was actually a study looking at it in experimental models following subarachnoid hemorrhage, which did show some relief of cerebral vasospasm with SPG stimulation. But where does the blood-brain barrier come into this? In 2004, there was a group out of Israel that first looked at this phenomenon. Essentially, after prolonged sessions of stimulation, they were measuring fluorescent tracer concentrations in CSF superfusate and found that these levels progressively increased over time with stimulation. And they later did the experiment in a handful, less than 10 canines, and saw very similar results. And since then, there really hasn't been much work done on the sphenopalatine ganglion. As a lot of us know, it's now a treatment for cluster headaches using SPG stimulation, so we felt that it was high time to sort of revisit this phenomenon and seeing if it's something that we could potentially harness. So we had some questions based off of this old data and some preliminary work in our lab. We wanted to see if we can develop a model to quantify this in rat-brain parenchyma, as the previous studies had been done in CSF. And we wanted to see what the dose response was in terms of the frequency of stimulation. The previous studies had been done only at 10 hertz, but we now know in human trials as well as in animal models that there is a very significant frequency-dependent response of SPG stimulation, where essentially low frequency heightens the parasympathetic interactions and high frequency suppresses those responses. So high frequency stimulation is typically what's used for treatment of cluster headache, the idea being that it's suppressing that parasympathetic response. And last, we wanted to see if there were any regional differences. We know that it does supply the territories of the anterior circulation, but we wanted to see if there was any rostrocodal or lateral distribution of blood-brain barrier modulation. So our protocol we developed, we catheterized rats through the femoral vein. We exposed the anterior ethmoidal nerve, which is a postganglionic fiber from the SPG, just as it enters the ethmoidal foramen right behind the eye, so a retro-orbital approach. You can sort of see our surgical setup there. We have a 90-second on period of stimulation followed by 60 seconds off, and that's one cycle. And we do a six-cycle priming session followed by a 10-cycle infusion session, where at the beginning of each cycle we inject 0.1 cc's of the fluorescent tracer and then inject another 0.1 cc's of heparinized saline to flush the catheter. And this is just the tracer we use. It's a 70 kilodalton fitzy conjugated dextran diluted in heparinized saline. And our populations, we have a control population of rats, sham surgery, which involves exposure of the nerve and hooking of the electrode around the nerve. And then we did 5-hertz stimulation, 10-hertz stimulation, and then high-frequency stimulation at 200 hertz. And something we wanted to investigate was continuous stimulation. So as I mentioned previously, the on-off cycles of 90 seconds on, 60 seconds off, what would happen if we just left it on the entire time, and if there was any difference in response. So after stimulation, the rats were immediately harvested. Their brains were sectioned into different regions, so 5 in each hemisphere, so 10 sections total. They were homogenized in trichloroacetic acid and then plated alongside a standard curve of concentrations, as well as a serum specimen. And our calculation that we were looking at was fluorescence uptake in the brain tissue, so the percentage of serum concentration that was in the brain. That way we could sort of standardize it across all of the rat specimens. So this is our results. We found control rats don't see much of a response at all, about 0.25% uptake. At 5-hertz, we see a slight increase, and then as we sort of suspected at 10-hertz, we see this maximal response. Not on this graph. We recently do have data at 20-hertz as well and find that that is the step-off point when we stop getting fluorescence uptake. And we see it up here, as you can see, at the 200-hertz. And interestingly in this data, we found that when you continuously stimulate, you lose the response. So from a mechanistic standpoint, that's something we're investigating a little bit further. And the other interesting point in this is that sham surgery actually did elicit some sort of response. So our thinking is the hooking of the electrode actually causes some aberrant firing. And when we look at people who've studied this previously, all of their sham surgeries just included exposure of the nerve and not actual engaging of the electrode. So we think that whatever low-frequency aberrant nerve firing could be occurring when that electrode is hooked, it's eliciting some sort of response. So after getting this data, sorry, this is the regional differences, and we actually found no difference in uptake that was statistically significant. We did see a trend in the ipsilateral hemisphere at the 10-hertz stimulated population, but it did not reach statistical significance. And this is also what we had sort of expected given that in rodents it does have bilateral innervation. So we see a clear dose response that peaks at 10-hertz stimulation, and it drops off with high-frequency stimulation as well as continuous stimulation. And we wanted to see what happens if we look at this under a microscope. So we did the same experimental procedure, harvested the rat brains and immediately fixed them in OCT. We did not, or sorry, placed them in OCT. We did not do any fixation. We did not do any staining to try and get a pure image of what was going on without risking washing out any of the tracer. And we standardized our imaging analysis of this to make sure that we weren't, you know, adjusting any backgrounds or anything like that and that what we were seeing was the actual true fluorescence. And we reviewed it the same day as mounting for the same reasons to limit any fading. So you can see in our control population on the top, there's some faint signal that's residual in the vessels. But for the most part, most of it is washed out. On the left side is a stitched image of serial 10X images. Left side is left, right side is right. And then we have a region of interest sort of just above the hemispheric fibers of the corpus callosum that we found where the optimal fluorescence was. So we compared apples to apples across specimens. And then the next image is sham surgery. And consistent with our quantification data, you can see that there is a slight bit of extravasation outside of the blood vessel into the parenchyma. And there is a pretty significant amount of stain that remains in the vessel, we think lodged in the vessel wall. And then this is at 10 hertz stimulation. You can see very robust extravasation of the tracer dye out to the rat parenchyma. So I think this is a really cool picture that really shows the phenomenon and sort of brings it home in addition to the data. And then when we do it at 200 hertz, it's very similar to controls. Just some trace filling within the vessel, but nothing leaking out into the parenchyma. So our conclusions, we do see a frequency dependent response. It's maximal with low frequency stimulation at 10 hertz. There is an effect at lower frequencies as well as with sham surgery. But at high frequency stimulation, the effect is lost entirely. And we don't see any regional differences within the anterior circulation. And these are some references and, of course, acknowledging my lab. And we do have some funding through the Neuroscience Institute as well as a charitable trust that gave us some money to work on this. So thank you. Any questions? We do have time for questions. John? Can you turn on the microphone? So this is actually full disclosure with what we've submitted now to the next big conference. We do see reversibility at four hours. So we essentially did the stimulation and then waited four hours and redid it. And we see that it closes back up. And why don't you think 10 hertz increases? So that, I think, is really interesting. So we have some hypotheses of what we think is going on from a mechanistic standpoint. We think something to do with creation of reactive nitrogen species in the vessel endothelium. And we're thinking maybe with continuous stimulation that you sort of deplete the required substances for that reaction very quickly. And so maybe there is a response, but we haven't checked different time courses, but maybe after 15 minutes of stimulation it's done and then everything washes out. Richard, I have a question about the surgical approach. It's an interesting microsurgical approach you did behind the orbit. Correct. How could you isolate the small SPG of the rodent from other things going through there, possibly trigeminal, maybe ocular fibers going back? Right. So there's a small, the ethmoidal foramen that is where the fibers actually enter the brain and where they provide the innervation to the anterior circulation. So there's a very clear, that anterior ethmoidal nerve is very clearly coming off of. The SPG in rodents is less of a discrete structure and more of this nexus on the actual periorbita. And so then they coalesce into the anterior ethmoidal nerve. You can see the nerve pretty, you know, we have an operating microscope, and you can pretty carefully dissect it off the tissues around there. But the stimulating electrode is a larger bipolar electrode. Correct. Was there any attempt to shield that to prevent spread of the current away from the desired tissue? Right. I mean, to the best of our abilities, but you're right, there could be some current spread to some of the surrounding tissues. Something to work with. Yeah. That's fascinating work. Additional questions? Do you think it would work to stimulate the hypothalamus directly? That's an interesting idea. It might. I don't know. I mean, I think that that would be, from a translational aspect of that, would be difficult to perform in terms of translating it into like a possible human model. And there would be probably some side effects from that that would need to be figured out. But that is an interesting idea, to just directly activate the parasympathetic. Yeah. Thank you.

blood-brain barrier modulation

sphenopalatine ganglion stimulation

cerebral blood flow

low-frequency stimulation

therapeutic interventions