Well, good afternoon. Welcome to the AANS section on tumors, skull-based surgery section. On behalf of Dr. Link, I'd like to welcome you to this session. We have an excellent lineup today. I would like to remind everybody to self-report and see me for live virtual events to my AANS. Dr. Link, any opening comments from you? No, I think we have a really exciting program lined up. The session is set up. We're going to have several talks by leaders in skull-based surgery that are going to cover a variety of approaches. Then we have some award-winning abstracts which will be presented at 3.30. We'll take a short beverage break, a 30-minute break. We're excited that everyone can join us today. We might as well get started. I'm really honored and pleased to introduce my co-moderator, Dr. Ian Dunn from the University of Oklahoma, who's going to discuss the orbital zygomatic approach. Ian, why don't you take it away and get us started. Also remind all the speakers, of course, to try and stay on time. Looking forward to the session. Dr. Link, thanks very much. I will promise to stay on time and greetings to all of you from Oklahoma. I'll go ahead and get things started. Really, the purpose of the talk is to really emphasize a modular set of approaches, the full extent of which I'll discuss with you is a COZ. The point of the talk, however, is not to compare a trial with an eyebrow and a nasal or other approaches with COZ. It's really to emphasize when you may wish to employ this modular overlapping series of approaches. The parting words, I think my view towards COZ has always been best encapsulated in the immortal words of Dr. Al-Mefdi, which is COZ, I love you, thee, and I will attribute that to him. The concepts really, not only to minimize brain retraction with appropriate bone removal, but really to unlock anatomical corridors. Micro-surgical principles form the basis of many of our approaches and their reconstructive possibilities as well. Just one slide about the debate about whether or not we need more than trial when we're doing a transparent approach and we still have those discussions today. I was fortunate to spend time with both these individuals and one would say no and one would ask me why we're discussing not using a COZ. It's a fascinating evolution of approaches. Just in two slides from Cushing's transfrontal and Dandy's pre-terrinal and Fraser's forehead approaches to the basis of Yasir Gill and Fox and others work demonstrating the more modern frontotemporal approaches, which included over the years, the addition of the superior orbit first and then extending that orbital osteotomy down to the lateral orbit, extending the frontal craniotomy down to the middle fossa to what we would consider today is a cranio-orbital zygomatic approaches. These two figures represent the two that we'll speak about today. Certainly, I want to acknowledge that there are many ways to do this and many authors have contributed to this literature over the years. In the interest of time, we'll review two main versions of that. One on the left from Ben Lovren's group, where the cranio-orbital flap and the zygomatic arch are removed in one piece, and the one on the right, this is the late Dan Piper who was the first author on this paper, is a cranio-orbital flap and the zygomatic arch is left on but it's osteotomized and we'll talk about that in turn. In truth, again, these are tailored approaches. The base unit for me at least is McCarty's keyhole, and the equator of that keyhole is the superior orbital rim. North of that, so frontal to the south is the periorbital. You can extend this superorbital flap immediately down to the floor to the malar eminence along the lateral rim, and you can incorporate either removal or reflection of the zygomatic arch in these approaches. This is just a tip of the hat to Colin McCarty from the Mayo who first described this keyhole really with surgery for thyroid disease. Some advantages of incorporating orbit or zygomatic arch at surgery, less frontal and temporal retraction, the ability to, as you remove bone inferiorly towards the orbit, the ability to look superiorly and also from a lateral to superior approach. Do you have to take the zygomatic arch to get to the floor? No, but an advantage is extending to the floor permits access to the middle fossa, permits a very wide fissure split. It has that temporal lobe being able to move out more laterally and you have middle, posterior, and trans-silvian access. You can access in a COZ the anterior, middle, posterior, and a fossa we don't often discuss, but certainly is accessible is the infratemporal fossa. Of course, you have orbital access and a nod to neurosurgeons to not ignore the orbit and not totally turn that over to the other specialties. There's still a lot we can do in the orbit as neurosurgeons. We have access to the paranasal sinuses where we often encounter disease and we have reconstructive ability. But a take-home is that these modular series approaches allow enhanced exposure coupled with microsurgical principles. Although it's the bone that we're talking about, there's much to consider in these approaches before the bone. Personally, I imagine the clinoid is perpendicular to the floor and I balance the desire for frontal over-traction, a relaxation rather, an extension. Balance that with the excessive clinoid ankylation where the clinoid is running away from you if you overextend. Of course, I like navigation, especially in the infratemporal fossa along the pterygoid plate. I'd like to be precise around where the frontal sinus is, the supraorbital nerve is an imperfect estimate of the lateral aspect of the frontal sinus. With the orbital rim off, it's easier to place neuromodulin leads for extraocular muscles as well. Always try and preserve the STA during the approach. A nod to our endonasal colleagues, we also have a rescue flap in transcranial surgery as all of you know. We don't take that incision down to the bone. We try and save that pericranial flap, undermine that posteriorly, and keep that intact by just mobilizing it immediately, and having that relaxed, and being able to use that if necessary during the procedure. Many ways to handle the muscle. I've highlighted the one that we'll talk about today, which is a sub-fascial approach incision, at least a centimeter behind the keyhole, behind the superficial fat pad, through the superficial and deep fascia above the muscle. Keeping that about a centimeter behind the keyhole will help avoid a stretch on the temporalis branch of the facial nerve. Respecting the deep temporal neural structures when you're mobilizing the muscle inferiorly, and the concept is to mobilize the muscle inferiorly. Scalpel goes forward, muscle goes inferiorly, keeping that triangular region free of muscle bone. We're going to skip the bone in orbit for a minute and just talk about some reconstructive principles with sinus entry. Of course, you're generating the sinus, perinealizing it, obliterating it. I only use autologous tissue. I don't use gel foam or wax or any of that. Again, you've got the pericranium available for vascularized tissue. I don't personally have a muscle cuff. If you've not used the pericranial layer, you can re-suture the muscle of the pericranium. Otherwise, you can make small bone channels and re-suspend the muscle on its anaphylactic position. Don't forget to replace that zygomatic arch. Of course, we like to consider aesthetics in these approaches as well. You can see that coronal here re-applies very nicely in that craniorbital flap. If you harvest some lightly irrigated bone chips from your burr hole placement, you can use that essentially bone grout or bone putty that results to fill in these small bone gaps, especially your frontal. Cosmesis is important as well. Let's talk about the first one version of the craniorbital zygomatic approach. I'll call this COZ in that it's a full removal of the craniorbital flap with the zygomatic arch together. The key is the zygomatic arch is removed. The second area on the right indicates the inferior orbital fissure is a key component, is that there's a connection between that and your Ricardi's burr hole. In this video we prepared for this lecture, it's going to go quickly through what that looks like. The key in the skin incision, it really goes close to the contralateral superior temporal line, harvesting the STA. Here's that subfascial dissection behind the keyhole, about behind the terion, is a subfascial dissection. If you take this dissection in subparastrofascial fashion, you're able to easily expose the lateral orbit of that zygomatic facial complex. Reflect that muscle inferiorly. Here's the Ricardi's keyhole, periorbita and frontal dura. This is a sped up version of the craniotomy, but you'll see the inferior orbital fissure being connected to a portion of your Ricardi's burr hole right there. Here's that malar eminence cut at the top of the screen, zygomatic osteotomy. The last part is, you're attached to the superior rim, and so the Ricardi's keyhole gives you a portal through which to use a chisel and take that osteotomy to remove the flap. The next is a different version, is the craniorbital plus the zygomatic, so it's craniorbital flap and a zygomatic osteotomy. And the difference being, the zygomatic arch is left attached to the mass and muscle. It's osteotomized anteriorly and posteriorly. Points of attachment are the medial orbital rim, the sphenoid lateral orbit, and the orbital roof. And once you remove that craniorbital flap, you can also remove the superior lateral orbit as well, although many people just drill, any of those protuberances very flush, and that works well too. And the osteoidal locations in that version, the zygomatic cuts anteriorly are along the line from the back of the lateral orbit, extending inferiorly and posteriorly at the root. The mass first stays attached, that the arch is floating and it's not removed. Again, it's attached to the mass center and the temporalis is reflected inferiorly over it. And the inferior lateral cut is really made along the line extending from the top of the zygomatic arch, just above the malar eminence, just taking the McCarty's keyhole. This is just a quick, the skin and muscle are as the prior video, just showing the zygomatic osteotomy anteriorly, posteriorly and anteriorly here. McCarty's keyhole outlining the craniotomy here in a second, this is sped up. And the last version of this is going to be that lateral orbital osteotomy. And this is a little higher than you would usually take it, but you take a little lower and you then, once you're through, you connect up to the McCarty's keyhole, you can remove that. That flap in the last point of attachment is that superior ribal rim, which you can osteotomize through the McCarty's keyhole. And once again, you can remove that superior lateral orbit as a separate piece. Once that's done, this is just the, of course the bone opening, is that you have so many possibilities of how to access different anatomy. You can take the mini-orbital band and create an extra extra-orbital section for a clinodectomy as well as lateral aspect, the lateral axis, the cavernous sinus. You have, once you open the dura, a number of corridors for paraceropathology. You can unlock that anatomy through clinoidectomy and open the rings. You can create space through debrilling tuberculum and create more room in front of the optic chiasm. You can take that clinoid and optic canal and really create a number of different working channels, work in the orbit, unlimited trans-soviet access, access to posterior circulation, laminate terminalis, middle fossa floor. You can access the posterior fossa from that route through Clostrid's triangle. Again, infratemporal fossa, cavernous sinus, and then reconstructive possibilities as well. These are all examples where we would do that. Again, mobilizing that clinoid, not for pathologic reasons, but to allow that carotid to move, expand your corridor, and that's just simplified through a bone removal through a CO and OZ approach. Reconstruction, I've entered a peroneal sinus here, small amount of muscle. In sparing that pericranial layer, you're able to sweep that down and vascularize that as well. And again, getting really down to that floor allows you to access the posterior fossa through the middle fossa if you need to do that, if you have a multi-compartmental tumor as well. A couple of cases, some may do this endonasally. If this was done cranially, COSD is a great option for that. This is a, I like to open the dura low frontally and low temporally, keeping the dura on the brain that we're not gonna do anything with, and make a relaxing cut along the salivary fissure. And this just highlights a few, a wide fissure split enabled by the COZ approach, laminar terminal analysis access, being able to remove a terminal of the posterior fossa, show the basilar complex here, expand that working channel through posterior plantar dectomy. I like to use a sonoped for that. And this reconstructs very well. And a couple of points here, you can see the zygomatic gasteotomy in the center and the nice cosmetic application of that flap after surgery. Excellent orbital access with the COZ, of course. And again, this is an 87-year-old actually with bractosis and left the zygomatic arch on, cranial orbital flap only here. Again, this is a modular overlapping series of approaches and you can, to minimize recurrence, have to take all that parotid sharply. Excellent orbital exposure with a cranial orbital flap here. And this shows a similar theme, but with something extra. With any of these tuner wing tumors, really need to inspect that middle fossa floor. So you can see your areas of computer recurrence would be along the parotid and look at all this infratemporal fossa involvement in the pterygides below the middle fossa floor. And left the rim on here, but again, took the arch down, really facilitated the ability to get down to the infratemporal fossa. Here we're taking, this is getting towards the infratemporal fossa, drilling out the entire middle fossa. This is cutting tumor off the pterygid muscles and dropping the zygomatic arch really facilitates that exposure. These are all recent cases from out here in Oklahoma. This is a 37-year-old professor with headache and dropping that zygomatic arch, getting very low is very helpful. So this is a traction suture. I learned that from Molly Christ, bringing that temperature up, getting very low on the floor. And the key to unlock this operation really was finding the right plane and getting very, very low. And then being able to work up, we have a very nice plane identified here in what turned out to be a schwannoma and we're able to get an excellent resection. But the COC was a beautiful approach for this, both for being able to look up and get very low to the floor. And last case I'll show here, it's a multi-compartmental meningioma, meningitis, MAVIS case now. This is a sphenoidal cavernous sinus, infratemporal fossa, muscle orbit, and left the rim on. You can see the video nerve here, infratemporal fossa involvement, deliver the tumor from the sinus around B2. Here's lateral cavernous sinus access and being able to explore the cavernous sinus rather. And you can see everything laid out here. Really nice to visualize that anatomy through a zygomatic approach. So in the last few seconds, all the considerations for these, it's really an overlapping series of approaches, a terminal aspect of which, a full monty of which is a COC. At the end of the day, it's still microsurgery. We think about the frontalis, we think about the pericranial flap, the zygomatic arch, the keyhole, the craniovertal flap, all the extradural work you might do, optic canal, orbit, clinoid, middle fossa, infratemporal fossa exposure. And then that's, of course, before you even do the surgery. So how you handle the olfactory tract is important. How wide the fissure split. Other fossa you're gonna enter, of course, reconstruction as well. Thank you very much. Okay, hello everyone. My name is Gabriel Zada and I will be discussing endoscopic approaches to the anterior skull base. I wanna thank Dr. Link, Dr. Dunn and AANS and my esteemed colleagues on this panel. Here are my disclosures. And I'm gonna try to cover the catalog of minimally invasive endoscopic techniques to expose and address anterior skull base pathology, highlight the benefits that endoscopy and keyhole MIS approaches provide and review the most commonly used approaches to facilitate access. It's no surprise whatsoever that the endoscope has contributed a lot and increasingly so to neurological surgery over the last several decades. It's really an amazing evolution with pioneers such as Dandy and Hopkins. And in the last several decades, we've been able to address skull base pathology with the first report of pure endoscopic transplant surgery now a quarter century ago. And the development of so many extended approaches over the last several years. On the left is the first endoscopic video I know of for a pituitary tumor by Professor Guillaume used as an assistance agent in 1962. And now in 2021, the extended approaches we can perform to address a wide spectrum of skull base pathology is pretty amazing. We can access really the entire skull base with minimally invasive approaches. And the endoscope is a major tool used for that. And so I'll be addressing mostly anterior skull base pathology. The catalog of neuroendoscopy really lends itself to a minimally invasive tumor service line with the majority being endonasal approaches, but a wide variety of other ones such as endoscopic assisted keyhole craniotomy which I'll be addressing as well. So to kind of dive into the endonasal approaches, there's really a wide spectrum going from frontal sinus pathology back along the olfactory groove with transcribriform approaches, transplanum approaches and transtuberculum approaches in addition to direct approaches to the pituitary gland. The direct approaches make up about two thirds of our approaches endonasally and the extended ones the other third. And we maintain the principles of skull base surgery. So for micro adenomas, we can obviously treat a wide variety of pituitary tumors. This is an example of a Cushing's micro adenoma and you get unparalleled views with the endoscope and able to perform really extra capsular dissections which we know are beneficial for Cushing's micro adenomas. And then even for larger pituitary tumors, we rarely require extended approaches. That's because they lend themselves to descending from the supercellar space and often have a wide aperture through the diaphragmacella. An angled endoscope can provide views looking upwards to make sure we achieve our goal of decompressing the optic nerves and chiasm. And also one reason we don't routinely require extended approaches for pituitary tumors. When possible, try to perform an extra capsular approach. I find that this is more useful for firm adenomas. And here you see preservation of the normal gland and a complete extirpation of a rather firm pituitary macro adenoma. In some cases, endonasal approaches may not be the way to go and there are risk factors for this. You just saw a beautiful talk by Dr. Dunn describing the modular benefits of the cranial OZ approaches. Tumor consistency is a very important characteristic that we're learning more about. We grade our pituitary tumors and meningiomas objectively and find that the firmness of the tumor and consistency is associated with extent of resection and intraoperative issues such as intraoperative CSF leak. The goal is to be able to predict consistency on MRI in the next several years. Extended approaches to the anterior skull base are used for a wide variety of lesions not accessible several decades ago with these approaches and mandating the craniotomy in the past. There are a variety of options when addressing anterior skull base pathology. As you heard, the terioneal OZ approaches, of course, the bifrontal approach, superorbital keyhole approaches, interhemispheric craniotomies, and of course, a variety of endoscopic and nasal approaches. You want to make sure you have all your instruments lined up and plan for reconstruction. When doing extended approaches and working around the carotid, you need to make sure that you have endovascular treatment available and all your instrumentation for bony removal and tumor removal as well. This was an epidermoid lesion of the cavernous sinus going back to Meikle's cave that was resected via transteragoid endonasal approach here. Just fast-forwarding to this to show the removal of the epidermoid contents, going all the way back to the posterior fossa there, and then reconstruction. Starting with some more basic anterior skull base pathology in the cribriform region. This was a patient sent to me after a surgeon attempted a pituitary tumor removal and had an iatrogenic defect in the anterior skull base causing a meningoencephalocele. This shows how extended approaches to the olfactory region can be used to address meningoencephalocele when they do arise. Most of these in this region are spontaneous, but you can have some iatrogenic ones as well. These are good lesions to start with when developing an extended endoscopic practice. Then of course, more complex lesions such as olfactory neuroblastomas with extension into the anterior skull base, often mandating an approach, including radiation chemotherapy. This is a patient we treated with a transcribriform approach. We did have enough nasal septum available free of tumor with negative margins to perform a flap here. Just to show ourselves working around the cribriform region, removing the crista galli, and then doing an aggressive dural resection. Then removing the intracranial component on block, dissecting the tumor away from the olfactory tracts and bulbs and the orbital frontal arteries. Then removing this component of the tumor completely and then of course, focusing on reconstruction here and hemostasis. This is use of fascia lata from the thigh for this purpose, and then a pedicle nasal septal flap. This patient had a complete resection, as you can see here, followed by chemo and radiation. Moving more posteriorly along the skull base, when performing transplant and transtuberculum approaches, the relevant anatomy here, of course, the pituitary gland, tuberculum sella, superior intercabinous sinus, and both the medial and lateral optical carotid recesses. Of course, the carotid arteries and optic nerves and the superior hypothesial arteries. This is a removal of a craniopharyngioma, a pretty standard one. We've done our supercellar opening and are now debulking the craniopharyngioma. Then we work on lateral micro dissection, careful dissection away from the superior hypothesial arteries, and more laterally from the PCOM arteries and oculomotor nerves. Then dissecting the tumor away from the third ventricle, getting this great view of the mammillary bodies. Then for this case, I performed a gasket seal reconstruction as described by Dr. Schwartz and Dr. Anand. You can see the extent of resection here with decompression of the optic apparatus. It's important to understand neurosurgical anatomy from above and below, when doing open and endonasal skull-based approaches. This is just a good example of relevant anatomy. This is a rodent view of the anterior clonoid process, optic nerve, carotid artery, and of course, the optic strut deep to this. Here's a view of the same anatomical regions endonasally. Having this 3D anatomical perspective is very critical when doing these approaches. I mentioned that pituitary adenomas rarely require extended approaches. Here's a patient with a large tumor spilling over the tuberculum, cell into the anterior cranial fossa, also lateral extension into the oculomotor cistern, causing a right third nerve palsy and vision loss. This was a case where we did do an extended approach. Just moving ahead here. First, I just open up the cellular component to debulk it and get a sense of what the consistency of the tumor will be like. Then we do our trans-tuberculum, posterior transplanum extension. We're now working in the supercellar space, careful two-handed dissection. Of course, pituitary tumors are reported to have a pseudocapsule. These can still be invasive and require careful dissection from the optic apparatus and from the carotid arteries. This patient had a dramatic recovery in vision and of the third nerve palsy, and the MRI showed a complete resection, as you see here. Here's another gasket seal reconstruction with a flap over that. He's doing well to date with no tumor progression or recurrence. Craniofringiomas almost seem like they were designed for this purpose for trans-tuberculum endonasal extended approaches. That's because the long axis of the lesion is often in line with the endonasal approach, and then of course, many of them are cystic or softer tumors. Just some examples of craniofringiomas. This is a trans-tuberculum approach. We're doing our supercellar midline opening and a French door dural opening here. Opening the arachnoid and then debulking the tumor and then careful dissection laterally and then superiorly away from the undersurface of the chiasm. Then making a decision of whether the stock can be spared or not. In many cases, it cannot be. Of course, when a gross total resection is achievable, it should be done even at the expense of the stock in most cases. Here we are sacrificing the stock inferiorly to get a complete on-block resection of the tumor. Moving on to even larger, more complex craniofringiomas, which can be treated with these approaches. These mandate often a trans-tuberculum, trans-planum, and sometimes transclival approach. Here we've done our opening with the trans-tuberculum, trans-planum approach. We're dissecting this large craniofringioma away from the chiasm. We've removed the dorsum cella and posterior clinoid processes and are dissecting the tumor away from the basilar apex, from the third nerve on the left side. Then of course, from the PCOM artery and thalamoperforating arteries. Then looking back and dissecting the capsule away from the mammillary bodies and third ventricle, and removing the tumor once all that dissection has been done. Then inspecting the cavity. With this much dead space, I like to use fat and then either a fascia latae or a synthetic alloderm reconstruction with or without a rigid buttress and a flap to cover that. You can see the extent of resection was complete in this case. For meningiomas, I'm very selective about these because I do feel that sometimes a craniotomy is favored. This was a perfect setup for a trans-tuberculum resection for this meningioma. The tumor does not have a lot of lateral extension. This is my colleague, Dr. Roble, doing a flap and the exposure. Then we've done the trans-tuberculum bony removal. We're now cauterizing the dura and the superior intercavernous sinus. A judicious use of a microdoppler on every single case is a great habit to develop. Now we're opening the supercellar dura, cauterizing the superior intercavernous sinus. Then we start to debulk the tumor. We get a sense of what the consistency is like. We obviously devascularize it early. We're now getting our first glimpse of CSF over the top of the tumor. We're using an ultrasonic aspirator to debulk it and dissecting it away sharply from the optic chiasm. The next step is to roll the tumor down and find the pituitary stock. Of course, sharp dissection is preferred. We want to preserve the pituitary stock and gland in this case. Now you're getting a view of the infundibulum here. We roll the tumor down off the diaphragmacella with preservation of the gland. Her vision got much better, of course, afterwards. Here's another more challenging case. There was more hyperostosis, a larger dural tail, and also some left-sided optic canal invasion here. We're doing our transtuberculum posterior transplanum approach, cauterizing the dura. This one also had a more fibrous consistency. As you can see here, it was a little more calcified. We're taking time to dissect over the gland and identify the optic apparatus here, and then a very careful two-handed dissection. We're now going to follow the tumor out into the left optic canal, make sure the undersurface of the nerve is decompressed. We're now working over the tumor to dissect it away from the ACOM complex and branches of the anterior cerebral arteries. Now you can see the left optic nerve decompressed. In these cases, I prefer not to use a rigid buttress when I have the optic nerves and carotid exposed, so I'll do usually a two-layer fascial reconstruction or synthetic reconstruction with a robust flap to cover it. Here's the post-operative scan. Now, three years later, preservation of the gland and stock and ample decompression of the optic nerves and chiasm with improvement in vision. The endoscopic intranasal approach is, as you can tell, very useful for addressing anterior skull-based pathology. A great alternative is a superorbital approach, whether it's done with an eyebrow incision or with a standard frontotemporal hairline incision or even an eyelid incision. This is my backup for endoscopic intranasal approaches. You heard a great lecture from Dr. Dunn on the COZ approach and just keeping in mind that terianal approaches are very lateral to medial approaches. Sometimes a more anterior to posterior trajectory is preferred to work between the optic nerves akin to what you get with an endonasal approach, and OZ does offer you that, of course. Here's a chapter by Dr. Dunn in our textbook explaining that. I do like to use an eyebrow incision, which does give me a similar trajectory. It's an excellent MIS alternative to endonasal approaches. This was a larger planum sphenoid alley and tuberculum sella meningioma. This was a great case for a right eyebrow approach. The benefits here are the reconstruction is minimal. Here we are drilling a McCarty keyhole, 3 by 2 centimeter craniotomy. We then drill out the undersurface of the orbital rim without removing the rim completely. We've now come under the gyrus rectus. We're decompressing the tumor. We're dissecting it from the skull base. This film is sped up a little, but we're dissecting away from the chiasm, right optic nerve, and then reaching over to dissect it away from the medial surface of the contralateral optic nerve. What I like about this approach is the patient can often go home post-op day 1 or 2 without any concerns for CSF leak, or rhinorrhea, and you can see the extent of resection is a complete removal. They don't have sinonasal morbidity as well via this approach. Here's just another example of that. Then these tend to heal well, especially in patients who are follicularly challenged. I'll use this occasionally in pituitary tumors that are very firm. Here's an example of one I couldn't remove endonasally, so I came back in with an eyebrow approach, or a planum sunodality meningioma that can be removed that way. Then finally, if I'm doing an eyebrow approach, occasionally we use the endoscope as an adjunct to use an angled lens to look under the optic nerve where there's a relative blind spot. This can just help determine whether there's additional tumor there. You have to be careful with this and make sure that the light of the light source and temperature does not cause any injury of the nerve and you're irrigating a lot. This can give you an additional view behind the nerve of the optic canal. A big believer in versatility and using our optical technology interchangeably to get the best customized approach for your patients. I hope I've conveyed to you there's a variety of endoscopic MIS approaches available to address anterior skull base pathology. Again, special thanks to the AANS for having me today. Thank you. Thank you so much for having me, inviting me to talk on AANS, Michael, due to the organizers. I'm going to talk about the endoscopic endonasal transcavernous approach. I subtitled this lecture, Method to Madness. You will see why. I actually learned that it's a method to madness from Evandro de Oliveira, who many of you know, because I saw he was able to do these amazing operations going into the cavernous sinus through transcanal approaches. His method actually was based on understanding anatomy really well, going from the lab to the OR. I really applied that into my practice. When I started having the privilege of working with my colleagues in Pittsburgh, and I started doing endonasal approaches, I focused very much on understanding this anatomy the best I could. We're looking at the paraclonal region, and we're starting to understand this clinoidal region from both endonasal and endoscopic and open approaches. We described this concept of the middle clinoidectomy early on, which is very important to access the cavernous sinus and paraclonal space. From there, we move on and we continue studying the cavernous sinus compartments, how to navigate within the cavernous sinus. This is a classification that complements NOS classification, which is basically a radiological classification. Ours is more based on surgical anatomy in a way that help us understand which structures we'll find in each compartment, and how to navigate around those safely. Superior compartment, and about the third nerve, interclinoidal ligament. Posterior compartment, the sixth nerve, adrenalus canal. Inferior compartment, the sixth nerve going more lateral in the cavernous sinus. Lateral compartment, lateral to the carotid, where we need to find the infralateral trunk, and the nerve is going to the supraorbital fissure. I'm going fast with this because this has been described and the papers are out there available for those with interest. But then we also started learning about the importance of understanding the walls of the cavernous sinus. And we started describing these medial and anterior wall of the cavernous sinus as the periosteal and meningial layers. And we started looking at, as we were going into the cavernous sinus, we were seeing these ligaments. And we studied these ligaments carefully in the laboratory. And we call them the paracetal ligament, ligaments. And there is actually a pattern of these ligaments. In fact, there is one that is very common, which is the carotidoclinoidal ligament. This is almost universal. That separates the cavernous sinus space below from the clinoidal space above. This ligament is the anchoring, the major anchoring of the medial wall of the cavernous sinus to the carotid. And it belongs or forms part of the proximal dural ring. And just behind it, we have the interclinoidal ligament. In addition, we also described this inferior paracetal ligament, such an important one because it's the first one we find when we open them through all the cavernous sinus. And just behind it, we have the inferior hypovasial artery, which is one of our fears when we go into the cavernous sinus to remove tumors from there. We've also been investigating the anatomy of the clinoidal space because it's not easy to understand. You know, we are all used to the anatomy of the lateral clinoidal space, the one we expose with a clinoidectomy. We can call it the ventral or lateral clinoidal space. We're now also interested on the dorsal or medial clinoidal space, the one we see from an Indonesian perspective, medial to the carotid artery. And in this nice illustration that I've worked with Josh Klain, it summarizes a lot of this anatomy. And I realize now how important it is to work with illustrators because you can summarize the findings of your research in just one picture if it's well done. And this shows, you know, for example, the middle clinoid here underneath the CCL here, proximal dural ring, and then behind it, the interclinoidal ligament above is clinoidal space, the dorsal one, and below is cavernous sinus where you have the inferior parasitic ligament, the inferior hypophysial artery, the carotid, you have superior compartment, you have posterior compartment, you have inferior compartment. There is so much anatomy in this picture, but we need to have this picture in mind when we're doing surgery in the cavernous sinus. This is even more complicated. This is looking at it from posterior-lateral. You have to see the connection between the ventral and dorsal clinoidal spaces and the anatomy within the rings as we see it both from open but also from endonasal because there are tumors that actually do embed into this clinoidal space endonasally, and we now can actually access it and remove tumors from these areas. Based on this knowledge, we describe the step-by-step technique of the middle wall of the cavernous sinus. Actually, Professor Oldfield, the great Oldfield, described first the incidence of ACTH adenomas embedding the middle wall of the cavernous sinus and the technique he used to remove them. I actually watched him doing this when I was in UVA for a year, and it was fascinating, but it was so limited because of the macroscopic view. So our technique has expanded, incorporating all this new knowledge and taking advantage of the endoscopic endonasal approach which provides this much better visualization. So in this step-by-step, we can see how there is, for example, tumor remnant in the middle wall of the cavernous sinus. We can separate one middle wall from the anterior wall. It's so important to have the right instruments for this technique. This right-angle knife is key for this so we can safely open the anterior wall of the cavernous sinus. It's so important to really open it well so you see the carotid artery. And the difficulty of this is, of course, you have venous bleeding that you need to control and, you know, be comfortable with that so you can actually see the carotid. If I don't see the carotid, I don't feel good when I'm doing this operation. Once I see the carotid, I feel better. I can study and defend, for example, the inferior parathyroid ligament. You can coagulate, you can transect the inferior hypovascular artery just behind. Then we cut the tumor along the posterior clinoid process to detach it posteriorly and medially. And then finally, we detach superiorly the carotid clinoid ligament, depending on the extent, from superficial to deep. And finally, the whole middle wall can be removed and blocked. At the end, you have this picture, the carotid all exposed, posterior clinoid, interclinoid ligament. How do we apply this technique? For example, this young patient, 15 years old, with this devastating Cushing's disease, if you look at the imaging, there is tumor on the corner of the cell. Actually, my fellow was doing this case. He removed the tumor and actually it was positive, but I wasn't convinced because we moved tumor up to the middle wall. So I decided to open the middle wall of the cavernous sinus, and you see what I found there. A lot of tumor. Thickened middle wall with tumor inside. This tumor is growing within the ligaments and within the wall of the cavernous sinus. Now, since now I understand the ligaments and the attachments, I can go gradually and detach these ligaments, coagulate them one by one. You can see the inferior hypoglycemic artery going up there. I can coagulate it. I can cut it. I can remove this along the postural canal process. And finally, I can remove this in block. And this makes a huge difference in the outcome of this patient. This is a complete tumor resection. Patient's cortisol crashes the next morning. Cease and remission for long term and very likely cure for long. Similar case. This is the story of Ali. Very interesting story. And all this with his permission. But this kid had two operations before in a different country, and they consulted with multiple surgeons, and many of them said, or all of them said, there is very low chance for cure. But you look at the MRI, the tumor is just at the middle wall. So I was lucky to treat him, and he came here to Stanford. We accessed the lesion. The first thing is good exposure to expose them to the wall of the cavernous sinus. Once I opened the wall of the cavernous, there was nothing on the cell. When I opened the cavernous sinus wall, that is a tumor right on. I knew it was there. Now it's stuck to the carotid artery, and this is not an easy task to do. You need to very carefully separate from the carotid. First, I start around it. I separate immediately along the postural planet. Then I have to see the carotid. And as I see the carotid, I can carefully separate. And you can identify those ligaments again. So you can see there is a method to this technique. These ligaments are transected one by one. I find those attachments. I always worry very much about the inferior hypovasial artery. I have the carotid artery under constant visualization. I use a Doppler compulsively to make sure I always understand what I'm doing. You can see that very robust inferior parasympathetic ligament there. And I finally can remove this tumor completely and block away from the carotid artery, as you see right there. And again, this makes a tremendous impact in this patient that is now in remission and likely cured for long. Often they ask me, do you always remove the middle wall? And of course not. This is a case of acromegaly. I'm going to show you a few acromegaly cases. You see the middle wall in this case is pristine. It's completely intact. I can use my putty and clean the middle wall, and I don't see any area that is suspicious for invasion. So in this case, I'm not going to remove the middle wall. There is no point. But there are other cases. These tumors, for example, I suspect on the MRI already there is invasion on the middle wall of the cavernous sinus. And the exposure, you can see the tumor tracing down in that area. But look at the exposure. It includes both carotids. The whole cavernous sinus wall has been exposed. There is no way you can do it without the appropriate exposure. I try actually to remove the tumor off the middle wall in this left side, and the tumor actually is inseparable from the middle wall. So in this case, I have to actually remove the middle wall. That side is easy. This other side is a bit more difficult. Remember, acromegaly patients, they have very tortuous, big carotids. So it's always, you know, risky and concerning. We have to separate tumor directly off this carotid arteries. But here we are slowly making progress. You see, again, the infrared hypovisual artery. I'm going to coagulate that again. And that always gives me a lot of confidence that I can continue detaching the tumor from the carotid. And now I'm getting all the way up. That is the carotidoclonal ligament that I'm detaching and transecting it at that level and removing the middle wall and block. And you can see the whole carotid exposed. So similarly, this patient with this operation had a complete remission for long term. There are other cases that are even more difficult. And these are cases that invade into the cavernous sinus, but AcuRed doesn't do anything. Saxons don't do anything. These are fibrous tumors, and they're embedded in the middle wall of the cavernous sinus and stuck against the carotid artery. So you need to gradually detach them, find those ligaments, find the arteries, use a Doppler, and use the right tools. And with meticulous technique, being careful, being methodical about it, you can gradually detach this tumor. And these tumors that were thought to be incurable in the past, we can now actually cure them. This is NOS3B, where the cure rates are quite low. You can see the whole carotid skeletonized, and the tumor has been completely resected. This patient, two months already in remission. Even NOS4 cases, we can actually operate and enter into remission. Sometimes this is even easier because they are a soft tumor that you can resect around the carotid artery even easier than those fibrous tumors stuck to the carotid artery. So this patient also normal AGF1, not at three, but at six months. It took a bit longer, but normalized. So we learned that somatotrophic adenomas invade the middle wall much more often than any other tumor types. And this is now statistically significant and will be published soon. We've also realized that NOS4 grades are not that different, behave differently for growth of adenomas to the point that grade one, for example, more than half invade the middle wall, which seems to be very surprising, but that's what we found. And grade two universally invade the middle wall in our series. Our results have been remarkable for acromegaly patients using this technique over the last three years. We have over 90% remission rates just with surgery and all of our patients are in remission. The few with are not in remission with surgery are with either radiation or medication. Even I'm starting to do more and more cases that are being referred because the tumor has been left in the cavernous sinus and now the patients are facing the dilemma of whether to do radiation. Sometimes these are younger patients, females that are going to have reproduction, etc. Or you want to do medication with the long-term side effects, etc. So taking some of these patients into surgery and we can get complete remission of tumors that are left in the cavernous sinus. We can go after them and remove them. This other case where only the central part was the bulk, somewhere else I was able to take back and do a complete resection even it was a NOS4 on this right cavernous sinus. And now the patient is also in remission with no diplopia, no side effects. Because of course the question is what is the cost of this trans-cavernous surgery? And in our experience, this is two years not three, but still the results are similar. I didn't have any injury to the coronary artery, which is always our main concern, but with careful technique you can do this without an injury. And our rate of double vision has become very, very small and it's always been transient. Very few patients get double vision post-op and those that get it because of fascicular palsy is transient and gets better. So there's no permanent morbidity in this technique, in my experience. We can also apply this and translate it into other pathology, not just adenomas, but also for example chordomas, chondroit tumors. We described years ago in Pittsburgh this trans-cavernous interdural posterior canalidectomy. And you know, many years before Evandro had described this similar posterior canalidectomy doing transcranial, but now we describe it going endonasal, in a reverse way. And this actually makes a big difference in the impact on the outcome of patients, big impact, because as you know, you can see that beautiful ligament right there, inferior parietal ligament, we can transect it and once we cut it, we find inferior alveoli, we coagulate it. When we have a big client that we can remove completely. Getting a complete tumor removal, even a supra-total tumor removal in chordomas is critical, is super, is so important because these patients have a much better outcome that way. Even tumors that have wide cavernous sinus invasion, we can open the cavernous sinus and the same as with adenomas, we can remove chordomas from this area. In this case, we're looking into durellus canal, into the posterior compartment of cavernous sinus, but now we need to directly open into the cavernous sinus to get tumor from that area. And this technique can be learned and can be applied. This actually is my fellow doing the operation, not me. And this is his video and we are just resecting the dura around it. So there is a method to this madness of going to the cavernous sinus, but it is important to be very aware of the learning curve. It is very steep. It requires intense training and dedication. And it reminds me for those of you that visit Yosemite, like going up Yosemite, some of these cases look like that, but you need to prepare like Alex Honnold, you all know, did. Alex Honnold, you know, was able to climb a free hand technique, solo technique in less than four hours. The reason I spent so much time preparing this is because I was so scared, he said. But with enough preparation, it wasn't scary anymore. I knew I could do it. And these words resembled to me the way I feel when I go into the cavernous sinus to remove tumors that are stuck to the carotid artery. It seems impossible, but it's all about the training and the preparation together and to be able to do it. But remember, there is a tremendous learning curve. You need to respect that, but it can be accomplished. And I hope that I've been able to translate to you, you know, some of the anatomical descriptions we've performed, technical nuances, and how we can improve the outcome of patients. And through this be madness, there is method in it. Thank you very much. All right. Thank you very much. Fantastic talk. I know I've learned an incredible amount from Dr. Fernandez-Miranda. I was lucky to work with him and always learn something from that cavernous sinus talk. I'll be talking about transclival approaches, which does have some relevance and some relationship to even the transcavernous. These are my disclosures, which are not really relevant for this talk. Endoscopic endonasal approaches, of course, we've heard about them. And as we apply the approach, it's always done with two surgeons working side by side, which gives me the ability to work with bimanual dissection to accomplish all the things that we see being accomplished in these videos. There are multiple different models of approach. We heard about the transcranial anterior transcribiform approaches. We heard about transcavernous approaches, and now I'm going to be talking about transclival. The transclival approach is a very, very direct approach. That's one of the beauties of it. One of the advantages of this approach is it gives us very direct access to the whole region that we're trying to work on. And this is really where the advantage comes from that in the visualization. There's decreased cranial nerve morbidity if we address a midline tumor by coming directly at it. But of course, there are anatomic limitations, which I'll talk about being the carotid and the neural foramen up. There's a higher risk of postoperative leak in the posterior fossa. And there aren't great options for vascular control, but I'll talk about that in a bit. As I mentioned, this is a direct midline approach, a very straightforward access to this area. But of course, there are tips and tricks to be able to do it safely and as effectively as possible. The carotid artery, of course, is one of our lateral limitations. When we look directly at the clivus, this really is what limits our access. The clivus traditionally is divided into thirds. This is a drawing from Shaker's book, which shows the thirds of the clivus, which are broken down basically on approach with an orbitozygomatic approach being used for the superior clivus, a transpetrosal for the mid clivus, and a far or extreme lateral for the inferior clivus, or the inferior third. We can have a similar division from an anterior approach, but in my opinion, it's much more straightforward. It's all just adjustments of the same approach, but we can access anywhere from the cella up to the posterior clinoids, as we've heard about, down to the mid clivus, which is about the sphenoid rostrum, and then down to the foramen magnum. Looking at the upper third of the clivus, this is essentially what sits around and behind the cella. The most upper most portion of the clivus really is the floor of the cella, and here's an infracellar approach where we now approach all Rathke's cleftsys from below. Remember, Rathke's cleftsys occur in the pars intermedia, so they occur between the anterior and posterior glands, so rather than splitting the anterior gland, we remove the floor of the cella and open directly onto the cyst itself. This allows us to aspirate, it allows us the maximal resection of the cyst wall with no damage to the pituitary, and actually allows us to then leave an opening where we can marsupialize the Rathke's cleftsys into the sphenoid. So here's a little more inspissated contents being removed, and then finally looking up into the Rathke's cleftsys, here we have a beautiful view, and I know I can resect right up to the posterior gland posteriorly and right up to the anterior gland anteriorly to create the widest opening possible. This is a great example of marsupialization. Here you see films over a five-year period where, yes, there is still continued secretion of these contents, but rather than them filling the cella and causing symptoms, they can just drain into the sphenoid and not recur symptomatically. Another great example of that marsupialization not being symptomatic. As soon as we want to apply this, though, for other tumors extending behind the pituitary itself, we have to do the pituitary transposition that you heard Dr. Fernandez-Miranda describe, and really he's the one who came up with the current technique that we use. The initial pituitary transposition really was a direct transcavernous approach, but we would move the pituitary gland. This is what Kasam called the pituitary ligaments, which are the fibrous connections between the gland and the medial wall. Rather than going through the cavernous sinus, the gland was simply lifted out of the cella. Now this had the downside of devascularizing the gland and creating much more manipulation, but it's what taught us how that we can get to the posterior clinoid. So then I got where I would do these extra durally, where I would sort of peel the gland down extra durally and end up peeling tumor out, but this is really not nearly as safe or as controlled of a maneuver, and for a tall clinoid it can be somewhat risky or almost impossible. The inter dural or transcavernous transposition that Dr. Fernandez-Miranda described, both in the literature as well as previously today, is how we figured out how to do this in I think the most elegant way possible. And this essentially leaves the pituitary gland in its dural bag. We sacrifice the inferior apophyseal or try to dissect it, and it gives us beautiful access to the posterior clinoids. Here's an example of using that for a petroclival tumor by removing the pituitary, by doing the pituitary transposition, removing the posterior clinoid, we have access to almost all of this tumor. Here's another example of that being used. Here the cavernous has been open, we're coagulating the inferior hypothesial artery, and then once that's cut we then have access to the posterior clinoid. You can see the posterior clinoid being removed, and then once that's done we can get this kind of a resection of the tumor. So it really greatly advances our ability to access these tumors that extend up behind the cell itself. You might say, well is this necessary for some of these tumors? Well we looked at our chordomas and found that probably for chordomas I would argue this is critical. You know a wide bony margin and radical resection is probably the most important thing for chordoma, and we found that 70 percent of our tumors had histological invasion of the posterior clinoid. That's especially true for upper clival and petroclival tumors, so I would strongly suggest that a pituitary transposition is critical for any of these tumors. Extending now down below the cella to the middle third of the clivus, this is the most straightforward and probably the most familiar to most of us in where these approaches were started. It's important though to understand what the limitations and the critical structures here are. Here you can see the mid-basilar trunk, but more importantly the sixth nerve. This is really our lateral access and understanding this anatomy of the sixth nerve. Here you can see it curving up over the petrosal process of the sphenoid bone to go behind the carotid artery. Understanding this is what allows us to do safe resection. We can access even intradural tumors in this region, but before we go intradural you have to understand the structures between us and the intradural space. Here is the venous plexus, the basilar plexus, which is an intradural plexus. It occurs between the two layers of dura, between the periosteal and the meningeal layer of dura, and in fact this is where most chordomas grow. We think of them as being bony tumors, but indeed they spread in this intradural space. That's how they spread out towards Dorello's canal as well, and then once we remove the dura we have all the intracranial structures from an anterior perspective. Here's a relatively straightforward neurenteric cyst causing diplopia and vertigo and early days with a nasal septal flap, sphenoidotomy, and then simply using a bovie cautery. Now we would use a needle tip bovie and do an inverted rhinopharyngeal or RP flap, but you can see this very direct transclival approach respecting the paraclival carotid arteries staying below the cella, staying just at the level of the rostrum, and we have beautiful access directly in the mid-clivus. Now opening off to the side in this case because of the basilar being in the midline, we encounter brisk bleeding from the basilar plexus as it extends into the inferior petrosal sinus. That can be packed off relatively easily, and opening now with navigation directly on to the intradural cyst. Here are typical contents of a neurenteric cyst, and as soon as we make this small opening into the dura and into the cyst, we can see the basilar medially at the vertebro-basilar junction, and of course the sixth nerve right where it takes off. So these are the critical pieces of anatomy at this location. Here's looking out laterally towards the fifth nerve and the 7-8 complex. Of course this can be applied to much more complicated tumors, and in my opinion has become the workhorse and the preferred approach for clival chordomas. Here's a typical clival chordoma in a middle-aged woman, and again the concept here is to get the most radical resection possible. Here's a mid-clival tumor sitting between the paraclival carotid arteries, but we're getting a wide bony margin inferiorly, so the inferior bone has been drilled almost down to the foramen magnum. Then the tumor might be debulked just so we have better access to the bone. Here we're using a cartouche electrified stimulator to stimulate and look for sixth nerve stimulation, and then an important step of doing that inter-dural pituitary transposition. So we did our lower clival bony margin. Here's the upper clival bony margin. By doing pituitary transposition, sacrificing or dissecting the inferior hypothesial artery, we can then remove these posterior clinoids, and you really see how the tumor goes right up to the posterior clinoid, and this is necessary for a bony margin. Here we've preserved that inferior hypothesial. On the other side, I believe we sacrificed it. Once the bony margins have been achieved, we then peel this inter-dural tumor. Here you see it sitting between the periosteal, and believe it or not, even though this appeared to be intra-dural, it only touches or invades the meningeal layer. I believe that if it's invading this layer, you need to remove it, and you shouldn't be afraid of doing an intra-dural surgery. The best chance for long-term tumor-free survival for these patients is radical removal, and that includes resection of the dura. Here's our view coming up as we get a complete resection. We can see the sixth nerve coming into view. You can see a bit of tumor here down below in the inter-dural plane. Again, good reason for radical removal. We're stimulating the sixth nerve to make sure we don't damage it, and here's our final view from the vertebral basilar junction up to the basilar apex. We can see the left sixth nerve perfectly, and really a wide resection of the tumor. Here's our post-operative MRI, and again, looking at our series now, close to about 300 surgeries. Our gross total resection rate overall was about 58 percent in this early series, but one thing that you'll notice is, of course, that's better in primary tumors. Of course, this approach cannot be used by itself, but should be used in combination whenever necessary with an open approach to achieve radical removal. the things that we did notice is that larger tumors were more difficult, but we got better at this with time. This is the kind of thing Dr. Fernandez was talking about, going back to the laboratory and improving our surgical technique. Here you can see our learning curve, even in the first 60 patients, to get to the point now where we are able to achieve a complete resection in about 90% of patients across the board. Here you can see the complication rates, and I want you to pay attention to the much higher rate of carotid injury for these tumors than you might see with pituitary tumor. And that certainly has to do with trying to take the resection as wide as possible. There were no mortalities. Of course, we get recurrence of these tumors. Here you see a nice radical resection. And the thing we started to learn over time is where we didn't do as well was low and lateral, and that sort of makes sense. It's not right in the wheelhouse of that mid-clivus. So we started trying to understand and expand the lower clivus better. And what we realized is below the paraclival carotid arteries, we have a much wider ability to access. And this is where it was born, what I jokingly, but also I think makes sense to call the far medial approach. And we're gonna hear from skull-based master, Jacques Morkos, next about the far lateral approach, but these are very complementary approaches. They give us access to the jugular tubercle, whether from a medial approach or from a lateral approach, and being able to use them interchangeably gives us a much wider access to this region. Here's an example of a very extensive chordoma. And in this case, my only limitation becomes the hypoglossal nerve and the hypoglossal canal. We'll see here invasion of the jugular tubercle right up to the inferior petrosal sinus. So to be able to resect that, here I'm drilling and removing bone. As soon as I see that bleeding from the inferior petrosal sinus, I know I've achieved my resection. I don't wanna go beyond that because then I'll run into the pars nervosa. And by doing this kind of a removal, I can drill out the, you can see both jugular tubercles on either side. Both hypoglossal canals have been completely skeletonized and have done a radical dural resection. This is the kind of removal we can get through a far medial approach. So I really think this has become the preferred approach for tumors like chordoma. You might ask, if you're doing this condyle and even jugular tubercle resection, does this cause instability? Well, this is a study done partly with Dr. Prevodello's group, along with Curtis Dickman, who did the original far lateral stability. And they found a biomechanical inflection point at about 75% condyle resection. Well, we looked at this clinically and ironically, or perhaps very conveniently, we found the exact same inflection point. So that was 75% condyle resection. You know, you require a fixation, but I would argue probably between 50 and 75%, where you need to still be very careful, but less than 50% resection is perfectly safe to not perform an occipital cervical fixation. Here's an example of other types of pathologies we might use this for. There's a very large petroclival meningioma, and a middle-aged man presenting with significant gait difficulties, even has hydrocephalus. I'll skip through this largely in the interest of time, which I think we're running short on. We did this in two stages. And you can see one of the issues with these is the sixth nerve is completely serpiginous through here. We identified the left sixth nerve, but the right sixth nerve required a little bit of luck. And of course, some careful micro dissection to be able to preserve it. You can see sharply dissecting that nerve free. And here's our final view after resection of that tumor. We can check our vascular chart to make sure that's all stable. And of course, check our flap at the end. But we're able to resect all kinds of tumors, not just chordomas and not just extradural tumors. What about aneurysm surgery? This is certainly on the edge of sanity of what we should even be considering. Certainly for some tumors, I do think it's a reasonable option, but only if you've really achieved a quite dramatic amount of skill and practice, not just by yourself, but with your entire surgical team to be able to clip aneurysm such as this one. Talking about that far medial approach, here's a pica aneurysm that we treated. And this is an unusual pica aneurysm because unlike most which are lateral and perhaps exposed below or above the lower cranial nerves, here's a tumor, a ruptured aneurysm that was actually quite medial. So very anterior and medial aneurysm. So we approached it endonasally. Again, did this far medial approach where we exposed the jugular tubercle. And that's what allowed us to have this kind of access to clip this ruptured pica aneurysm. Even with an intra-op rupture, we're able to trap the vertebral artery and get a successful clipping of this aneurysm. Collibal reconstruction is always a concern. I'll give you the long and the short of this is that we've learned to use multilayer reconstruction to avoid this type of thing like encephalocele. And our dural reconstruction now consists of a multiple layers, which is an inlay with an onlay of fascia followed by fat graft to buttress against the CSF. And then finally a vascularized flap. Lumbar drains are critical. This has been proven in a randomized controlled trial for posterior fossa dural defects. You must use a lumbar drain. And conclusion, I think that this is an ideal approach for many midline posterior fossa lesions. We have direct access. Of course, we have neurovascular limits. There's a higher risk of postoperative CSF leak, but that can be managed and you have to learn and understand your options for vascular control. Golden rule in the end is don't cross nerves. If you have to cross a nerve, you should consider a different approach and you want to have the full spectrum of approaches available, whether endonasal or open approaches. Has to be bimanual dissection, must be microsurgical technique and must be multi-layer vascularized reconstruction. Pay attention to your own learning curve. And of course, I want to thank the AANS for inviting me to take part in this and invite everyone to join this free educational website, which is really a skull-based community designed for all. Thank you very much. Thank you. Hi, good afternoon. Ian and Mike, thank you very much for the invite. I trust you are seeing my slides, so I'm gonna get going. So I'm gonna discuss the far lateral approach and its variance. Let's go back to basics. The osteology and the roten anatomy. Of course, this is a region we need to be alert of. This is, you remember how roten used to put the pieces of bone as a puzzle and ask everybody to put it and recognize the various portions. So we need to be very familiarized with where the occipital condyles are. Again, to quote the great Al, they are at 10 o'clock and two o'clock. So they are interior to the transverse equator of the foramen magnum, which is why drilling the occipital condyle itself is almost never needed to achieve the intradural exposure. When you overlay the C1 vertebra on the condyle, that's what you have. And then remember that the hypoglossal nerve travels a supracondylar. It doesn't really travel through the occipital condyle. It is very hard to injure it as part of the drilling of a far lateral approach. Here is a depiction of the hypoglossal canal. And of course, a classic roten dissection from posteriorly. Here is a jugular foramen, which is important to understand if you are embarking on a far lateral approach and the relative relationship between the carotid and the jugular and the misnomers of vasa nervosa and of pars nervosa and pars vasculosa. Now, I am going to show you again, roten dissection showing muscle anatomy. And then I'm gonna show you dissections from our lab here in Miami as well. Not so much because you need to replicate this in the operating room. Because one of the rules is you really should not dissect the muscles separately when you do a far lateral approach. But you of course need to understand where they are and their names and so forth. So this is on the right side, semispinalis capitis, splenius capitis and sternocleidomastoid and then the trapezius is reflected. The second layer, it takes you to longissimus and semispinalis capitis is to your left. Notice the occipital artery has anatomical variants. It can go either deep or superficial to longissimus. Every specimen may be different. Then you get to the next layer of muscles which begins the suboccipital triangles. And of course, here they are, superior oblique, inferior oblique and then rectus capitis posterior major completes the triangle. Rectus capitis posterior minor has nothing to do with the suboccipital triangle. And of course, you can see where they insert from C1 transverse process to C2 spinous process. That's the inferior oblique and so forth. You can see posterior belly of digastric to your right. And of course, the occipital artery arising from the digastric ridge. Some schematics that we did to kind of simplify things, particularly when I teach and so forth. So this is a suboccipital triangle. Now back to Roten's dissection. So let's notice here where RCL is, rectus capitis lateralis is absolutely the lateral limit. If you are drilling or entering this muscle, you've just gone too far. It is unnecessary for the normal far lateral approach. Some Miami dissections in the surgical position using the hockey stick incision. Again, surgeon's view, the same muscles I just showed you on the right side, semispinalis capitis, splenius capitis, longissimus and the reflected sternomastoid. Again, this is not how you should do the actual surgery. Surgery, you keep everything in one piece as I'll come back to in a second. This is a suboccipital triangle to understand the V3 segment of the vertebral artery. You can see the occipital artery emerging from the digastric ridge. And again, a close-up view of the vertebral artery V3 after reflecting superior oblique. Why is the craniotomy called teardrop? Because this is a shape that you want because at the foramen magnum, the length of the craniotomy is less than it would be higher up. So that's why we talk about the teardrop craniotomy. There are tons of anatomical variations in this region. This is an extreme example of an extremely caudal loop of the pica. It's way down to C2, C3. We'll come back to some warnings about what to look for. Here we are in a cadaveric specimen cutting the first dentate ligament, which of course, make sure you don't confuse it for the spinal component of the spinal accessory nerve. It's much thicker, whiter, more glistening color. Of course, it inserts into the dura as opposed to the spinal accessory nerve. You may want to see the olive here. I'm putting one white dot on the olive. You can see it on your left. I don't think my pointer will show. So that's anatomy. What are the surgical variants? I want to start by saying that there is a far lateral, the transcondylar, supracondylar, paracondylar, and I will not discuss the very useful and unfortunately not well mastered by most neurosurgeons. It should be. It's a fantastic approach, the Bernard-Georges enterolateral approach that both cranial skull-based surgeons as well as spine surgeons, in my opinion, should become very familiarized with to approach the lateral aspect of the upper C-spine. So useful for chordomas of the throat and others. I'm not going to discuss it. I'm going to discuss the first four. And really this is mostly the work many years ago of Wen Zhu, one of Roten's fellows. So again, very briefly, I'm going to focus the rest of the talk later on the true far lateral approach, which is really the juxtacondylar approach. It is far lateral approach does not involve drilling the condyle. This is the far lateral. You can see the box in purple. That's the bone you drill after your craniotomy and you stop at the condyle. How about the transcondylar? Transcondylar has three variants. The first one is atlanto-occipital. I'm adding a shaded box on top of the bone that is drilled approximately for each of those variants. And I'm going to go fast because there is really no time to discuss each in detail. This is a complete transcondylar. And when you do that, you can see the hypoglossal nerve. And again, here is transposing the vertebral artery. How about the occipital transcondylar? This is occipital transcondylar. Now, how about the supracondylar? There are two variants. There is a transtubercular and there is a hypoglossal canal variant. Here is the transtubercular. This is a jugular tubercle, not to be confused with a jugular process. And I want to draw a parallel between jugular tubercle approach with what we're all familiar with, the craniorbital approach that Ian discussed in the first talk. I invite you to think about it. And there are immense parallels for each of the structures from blood supply to muscle, to outer bony obstacle and so forth. This slide summarizes the similarities between both. Of course, they're in different areas of the skull, but the concept is they're very, it's a twin of the craniorbital approach. Hypoglossal canal approach, this is the bone that you drill. And then you can see the hypoglossal nerve. Lastly, the paracondylar, that's both transjugular and transmastoid. Here is transjugular. And of course, transmastoid gets you into the posterior petrosa region. So those are the multiple variants of the so-called far lateral. The most useful one is the first one, is the juxtacondylar approach. How do I position my patients? I like four movements of the head. Spetzler had described three movements. I like to add that fourth one of translocation. You flex, you rotate, you bend, and you translocate upwards. This case illustrates the four movements, forward bending, lateral flexion, away rotation, and upward translocation to open up the OC1 joint. Roberto Heros had described many moons ago doing the lazy S incision. He actually changed his mind later on and used the hockey stick, which is what I also use currently and have always used. I use the black dotted line, but there are many variations, including the simpler blue line. However, it does not allow the exposure to be shallow when you do the paramedian straight line. What are the pitfalls? Preserve occipital artery. If you're doing a vascular case and you might need a bypass, preserve pericranium. These are the three groups of muscles that I discussed that you want to mobilize, but not separately. You take them essentially all as single myocutaneous flap. Use fish hooks rather than self-retaining retractor. Leave one centimeter cuff on the superior nuchal line so you can close the incision nicely at the end. Never use the BOV near C1. Beware of a very tortuous and ectatic V3 in elderly people and the vertebral venous plexus, which is quite easy to control. Variable need to remove bone. You need to remove the teardrop craniotomy and the C1 hemi arch, and then different parts of the bone based on the case you are doing. You need to be a tailor. So this is craniotomy, C1 hemilaminectomy, condylar exposure. And that's, in this case, condylar drilling, but the true far lateral does not include condylar drilling. Be aware of a completely closed canal around V3. Don't take your leg cell and bite off arch of C1 and take your vertebral artery with you. Beware of unusually prominent condylar emissary veins. This is an example of a C1 bony tunnel that can be very misleading if you're not careful. You will injure the vertebral artery. The dural work, I like to do a U-shaped dural flap, extensive arachnoid dissection, and use the various triangles to pass instruments. So illustrative cases for amen magnum meningioma, classic example. Here are the four positions of the head as I described earlier. Here is my hockey stick incision. And going quickly through the still images of the case, you can see it's very flat. Once I open the dura, I don't have any bony prominence in my way. Identify dentate ligament. You can see the vertebral artery pushed back by the tumor. Here is pica. You're working between all the triangles. Use every triangle you have. Start with devascularizing the tumor. Please preserve the radicular arteries that travel with the nerves. You can see C1 in the middle of the image. I actually put a temporary clip on it and the evoke potential changed. So I did not sacrifice C1. Be very aware that some radicular arteries are essential for spinal cord supply. And here we are halfway through the resection. And now we've completed the resection with the nerves intact, the vertebral artery intact, of course, and the tumor Simpson grade 2 resection. And that's a post-op. Perhaps you'd like to see more as a video. And again, another case, again, of a foramen magnum meningioma. I'm gonna show you a bigger case. You can see that the tumor crosses the midline to the other side. You see this tail to your left. It's very essential to drill up to the condyle to be able to do a cross-court reach. And you can see the vertebral artery partially engulfed in this tumor. So exposure is really essential here, good exposure. Again, same position, same incision. Here is the approach, drilling very thoroughly to flatten that acute angle. Now we can open the dura. And we are going, you see how flat the dura is? You can see spinal accessory. And we are going to mobilize every arachnoidal band that might be bothering us. We need to understand the blood supply to the spinal cord because we're gonna be working between those things. And we're going to use this non-stick bipolar to gradually, I usually start from the spinal caudal end of the lesion to gain some real estate to be able to see the tougher cranial part of the tumor. And that's what I'm doing here. and it's a matter, of course, as you all know, a matter of piecemeal resection with preservation of what's around us and working between the nerves and being very aggressive with dural resection. This is the final piece coming out, and of course, you're doing intra-op monitoring. At the end, it's a very nice view. The retractor was useful in this case to show me the top part of the lesion. Post-op imaging, it's a complete resection, nothing really residual at all that you can see on the MRI. Of course, we can have brain stem lesions that need far lateral approach, and I think I have time to probably show you that. I am going to speed through the video to show you the essential part of it. It was operated on elsewhere at a very good institution. You can see when it was operated on the left side, but it recurred, and now it's here. You might say you should go back to the floor of the fourth ventricle, but that is not the way to do it because the patient did have a facial palsy but incomplete. This is where the understanding of the anatomy as depicted by Roten and Ugur-Ture and Fernandez-Miranda of the brain stem fibers and the deep tracts. I chose the less obvious route, which is the far lateral approach, to take me round the bend. Here are the lower cranial nerves and doing the thorough arachnoidal dissection. Now we're stimulating to the middle cerebellar peduncle. It is a longer approach, but it is the one that physiologically makes more sense to preserve her facial nerve. Here we are removing the cavernoma. She did beautifully in spite of it being a redo case. You can see choosing the far lateral over the telovelar because of the facial nerve. Here she is six months later with no recurrence. Last, you might say when you push the far lateral more, you may do things like this, translocation of the vertebral artery. This is basilar invagination that normally I would have done through an endonasal approach. However, look how close the carotid arteries are. We did far lateral complete transcondylar approach. I'll show you still images of the case. This is a so-called ELTO approach, extreme lateral transodontoid approach. Still images from the case, left side mobilizing the vertebral artery, drilling the tip of C1, cutting the fascia around the vertebral artery at V3, unlocking it, putting a loop around it, lifting it out of its surgical bed, drilling occipit C1 joint and C1-C2 joint, and drilling the lateral mass, drilling occipital condyle, drilling the odontoid process, and that gets you all the way anterior to the cord, cutting transverse ligament, and my colleagues, spine surgeon, fuse the patient through the same operation. You can see the post-op. Far lateral is extremely important and useful approach, and I think I'm going to stop here, and I think you get a glimpse of how useful it is. Thank you very much. Thank you very much, Jacques, and to the rest of the speakers. The session was really great. A lot of tips and tricks and important anatomical information. We're going to take a short break now. We'll step off until 4 PM Eastern time, 3 PM Central, and then we'll resume. I'll talk about transtemporal approaches briefly, and then we have multiple award-winning abstracts from the tumor section that will be presented. Thank you once again to all of the speakers. We look forward to the second half of the session, and we'll go to break and see you in about 28 minutes. Welcome back, everybody, to the tumor section session for today at the AANS, the virtual AANS. I'm honored and delighted to have the opportunity to discuss a surgical technique topic I'm keenly interested in, and that is temporal bone-based approaches. I don't have anything to disclose. This is just a list of what I think are the most common or widely used approaches through and around the temporal bone to access pathology in the cerebellopontine angle, temporal bone and central skull base. Of course, everybody on the call is very well familiar with the retrosigmoid approach. Similarly, I think everybody is very familiar with a middle fossa approach or a subtemporal approach. And of course, this can be expanded by adding an anterior petrosectomy or drilling the bone of Coase's triangle, which provides an albeit fairly narrow but a good avenue to get from the middle fossa to the posterior fossa. The approach I'm really going to center on today, just based on limited time, is the posterior petrosectomy approach. All of these approaches I've mentioned so far are all potential hearing preservation approaches. You can theoretically do a partial labyrinthectomy, take off the posterior semicircular canal, and some people even think the superior semicircular canal, and preserve functional hearing in a patient. And that gives you some extra millimeters of room, which can be critical for some pathologies. Once again, I think everybody is quite familiar with the trans-labyrinthine approach to be used when either hearing is already lost or there's no hope of preserving hearing. The most extensive approach, of course, is the transcochlear approach, which honestly is rarely used anymore in my practice. That approach involves over-sewing the external ear canal and mobilizing the facial nerve out of the fallopian canal, and then the entire labyrinth and cochlea can be removed, which gives access right down to the side of the clivus. And Jacques has nicely discussed the far lateral approach right before the break. This is just a figure from Rob Jackler's Atlas on Temporal Bone Surgery, which kind of gives you the avenue and bone removed to allow access to the cerebellopontine angle and central skull base. I'll just point out that the retro-labyrinthine or posterior petrosal approach, it leaves this block of bone with the labyrinth, so it preserves hearing. The trans-labyrinthine approach gets rid of that, and taking out this triangle of bone, I will say, can make a heck of a difference in terms of the pathology that you might be dealing with. Now, often at our institution, working closely with my neuro-autology colleagues, we debate, should we go retro-sigmoid or posterior petrosal when we're operating on these lesions? And as you can see, the angle that you get is pretty similar, but there are distinct advantages and disadvantages. And when we have that debate at our institution, we ask a series of questions, either out loud or to ourselves, about what are we trying to accomplish here when we address different pathologies? So certainly one of the most obvious questions is, well, where is the pathology? For instance, this is a meningioma that's really in the middle fossa. I don't think there's any reason to do anything but a middle fossa sub-temporal approach to this. This tumor is all within the posterior fossa. While I suppose it could be accessed through a sub-temporal anterior petrosal approach, we would just do this through a retro-sigmoid approach, stay in the area of anatomy where the pathology actually is. It's important to think about where these lesions are in relation to the cranial nerves. So for instance, this tumor is behind and below the IAC. So I don't doubt everybody listening would just do a retro-sigmoid approach and cure this epidermoid. This is an epidermoid that is more in the petroclival region. I bet some people would do an anterior petrosal approach with a trajectory like this. I think it certainly could be operated retro-sigmoid. This is an example of a pathology that we really like to use a posterior petrosal approach for. Sometimes it matters if it's intraxial or extraxial. For something like this, a cavernoma in the brainstem, I really prefer a lateral approach through the temporal bone to be able to access the posterior cavity and make sure I clean out all the cavernoma. This is obviously an extraxial tumor. It can be accessed very well through a retro-sigmoid approach. You don't have to go digging around in the brainstem. Sometimes it's hard to tell. This turned out to be an exophytic ganglioglioma. Certainly the temporal bone can be the place where the primary pathology is. CT scan becomes very helpful to find out where the pathology is in relation to the carotid canal, the cochlea, the labyrinth, including the horizontal semicircular canal and so on. And that can help very much plan the trajectory. Even though a certain approach might seem good, you have to really look at the anatomy. For instance, this patient has this vestibular schwannoma, completely deaf, no word recognition. You think, well, let's do a trans-labyrinthine approach, but actually this would be a very bad head for a trans-labyrinthine approach. The sigmoid sinus is very anterior. This would be better accessed through a retro-sigmoid approach. Similarly, if you said, well, should we operate on this young woman with this small tumor and a little bit of hearing loss? This patient has a very high jugular bulb. It's kind of in the way, almost no matter how you approach it. So you have to be cognizant of that. But the most important question is, what problem am I solving by increasing the complexity of the operation? And I want to go through the steps in the posterior pterosal approach and why I think at times it is extremely useful. I'm going to first show you some cadaveric dissection pictures I've done here at Mayo Clinic in Rochester, and then I'll show you an actual case from an operative video. These dissections I'm going to show you were done by our Skull Base fellow from last year, Chris Graffio. He actually won the Balfour Award for Meritorious Research because of COVID. We had to have the award ceremony in my kitchen, but these next pictures are work done by Chris. And he was taught these techniques by my colleague and partner, Dr. Maria Paracelda, who joined us in January from Albany Medical Center and is now on faculty here in Rochester. So this is a left-sided approach. This is the incision that we use. These are 3D pictures, but unfortunately we couldn't figure out a way to display them in 3D for this presentation. We mark out where we think the transverse sigmoid sinus is. The scalp is reflected forward and inferiorly exposing the muscle. And we take the muscle in two or three separate cuffs that we can then reattach at the end of the operation. It exposes the squamous temporal bone, the occipital bone in the mastoid. We then place a series of burr holes. A lot of atlases will show placing a burr hole below the transverse sigmoid junction and a burr hole above the transverse sigmoid junction. I actually trough it out. I think it's much safer. Same thing with the transverse sinus. I trough it out. So we have the posterior fossa dura, the temporal dura, and then a couple more burr holes. So we can dissect the dura very nicely. This just shows, again, the transverse sigmoid junction and emissary vein, which you would be wise to control. And then the posterior fossa dura, middle fossa dura, transverse sinus, posterior fossa dura, middle fossa dura. And then we can make our craniotomy cuts very safely and elevate and remove this kind of L-shaped bone flap. All this takes about 45 minutes. And then working with our otology colleagues, they can come in and we start with a wide mastoidectomy, exposing the mastoid air cells. This is just a more close-up view. So as you work, all these air cells kind of coalesce into one main air cell at the top of the mastoid. That's the mastoid antrum. It's helpful to keep a thin rim of bone over the sigmoid sinus as you're drilling. And then you expose the antrum. And there's a couple important landmarks. One is the tip of the instrument here is actually on the incus. And this harder yellow cortical bone, that's the horizontal semicircular canal. That's a very important landmark because everything superficial to that horizontal semicircular canal, all that bone superficial to that has nothing important in it. The facial nerve is coming out at the inferior margin of the horizontal semicircular canal, and then turning down. The short process of the incus actually points to where the facial nerve is. You can tell at higher magnification, this hard cortical bone of the horizontal semicircular canal, you can actually see into the middle ear. In this particular specimen, the tympanic segment of the facial nerve was dehiscent. We could actually see it in there. And then the semicircular canals are skeletonized and not violated to preserve hearing. We don't usually do it like this. Dr. Graffio did an amazing job just to show the anatomic relationships. Of course, they're at right angles to each other. The horizontal canal bisects the posterior canal, and then the superior canal is the deepest canal in here. And then you can skeletonize the facial nerve. We leave it in its bony canal. We do not expose it to keep it protected, but we want to know exactly where it is. And then the remainder of the bone is taken away to expose the presymoid bone and the endolymphatic sac, which is being held up here by the dissector. This is just a more close-up view. So this is actually the window that we're going to work through to the posterior fossa. And so the next step is we open the dura parallel and just below the superior petrosal sinus, and then inferiorly just in front of the sigmoid sinus, and we over-sew the endolymphatic sac. There's some theoretical possibility if you accidentally stick your sucker in the endolymphatic sac, you can suck out all the endolymph and make the patient deaf. So we over-sew that, and then we can pull that dura up over the posterior semicircular canal. And this allows us a small window into the posterior fossa where we can open the arachnoid and get CSF out. Then the next step is we open the sub-temporal dura, looking out for the vena libae here. And then we can gently retract the temporal lobe all the way to the incisura. And depending on the pathology, you can often find the fourth cranial nerve. And then you want to cut the tentorium behind where the fourth cranial nerve enters the edge of the tent. That's shown here. And this step is really the key to this operation. This allows the sigmoid sinus to retract back and to open up this large space here out to the petroclival region. In a sort of far away view, if the patient, as in this specimen, has a lower jugular bulb, you can see the ninth nerve, some of the 10th nerve fascicles, the eighth nerve, the fifth nerve, and the fourth nerve. More close-up view again, the ninth nerve, eighth nerve, fifth nerve, fourth nerve, branches of the superior cerebellar and the anterior inferior cerebellar arteries. When we look deep inside there, we can see the sixth cranial nerve as it goes into Dorello's canal. And if we gently elevate the eighth nerve, we can see the facial nerve and the nervous intermedius and ICA running in between them. This is just a good example in my practice what we would use this procedure for. This is a 60-year-old woman from here in Minnesota who developed intractable left face pain consistent with trigeminal neuralgia. And you can see that she's got this moderately large tumor here. And it's about centered half above and half below the tentorium. And undoubtedly is stretching or compressing or pushing or possibly even invading her trigeminal nerve and causing her intractable face pain. And I, again, don't doubt some of the experts who have already spoken. And certainly many of you who are listening in the audience would say, I would operate that through an anterior petrosal approach or from an endonasal approach or even through just a standard retrosigmoid approach. You can see again, it's situated all anterior to the seventh eighth cranial nerve complex. It wraps around fairly far anterior heading up to the carotid artery. And this is an operation that we did through a posterior petrosal approach. So the craniotomy has been done. This is the left side. This is temporal dura. This is the mastoid. Decompressing the sigmoid sinus here. And so now basically all the bone drilling is done. This is the hard cortical bone of the inner ear. And I just wanted to show this. That this is one of the, let's say downsides or one of the things you often have to deal with when fully decompressing the sinus. So the facial nerve would be here. So you can see that we've got an opening right at the junction of the transverse sigmoid sinus and the superior petrosal sinus. And sometimes that happens and you have to deal with it. And I find that you can directly suture it. You can fix it. You have somebody slow down the flow by pressing on the proximal sigmoid or distal transverse sinus. And you just got to carefully place these 5-O-proline sutures to get this slowed down and stopped kind of one by one. You know, you got to alert anesthesia obviously that you're going to have a moderate amount of blood loss till you get this under control, but it is just venous bleeding. And eventually you can get it all controlled and tied off and preserve that sinus. I think it's very important obviously to always try and protect and preserve the sinus. So once we get that stopped, so now we have the sinus is okay here. Here's the labyrinth bone facial nerve heading down. So first we're going to open that dura just below the sinus here to get us into the posterior fossa. So we open up just below the superior petrosal sinus, and then we can cut down, and we're going to cut right across the endolymphatic sac, which I'll then just hold closed with my forceps. And then we'll tack that dura up. I'll just move ahead. And now we can look into the posterior fossa. We can see the eighth nerve. The seventh nerve has been pushed down. I can stimulate it right on the underside. This is actually the fifth nerve, which has been stretched way inferiorly by this tumor. Obviously, that's not a big enough opening that we'd want to operate this tumor. So now we've got to open up the temporal dura. And we extend kind of right down to the superior petrosal sinus. And we've got to be careful here of our venal lobe. And now we're going to clip, ligate the superior petrosal sinus. And then we can finish opening the temporal dura and elevate the temporal lobe. So now we can look out for our tentorial incisura. So I just gently put a telphapatty on there. And I'm looking for the edge of the tent. And then we can go ahead and cut the tentorium, which is also a real bloody undertaking. There's kind of the final bit of it. And so now we're actually going to have a big window into this tumor that extended both supratentorial and infratentorial. And we can resect the tumor. This is just the pre and post imaging. You can see there's always a little bit of tumor I leave because the sixth nerve is entering Dorello's canal right there. I've decompressed the fifth nerve. I leave a little tumor in Meckle's cave. And we're able to get that supratentorial and anterior part out through our mental fossa exposure. And I'm going to stop there so we can move on. We have many award-winning abstracts. So we'll proceed to those. Thank you very much. Okay. Well, thank you very much, Dr. Link, for that fantastic presentation, as well as to all the other masters of skull base surgery. They're going to be hard acts to follow. My name is Dan Eichberg, and I'm one of the neurosurvey residents from the University of Miami. And today I'm going to be discussing our experience with stimulated remodestology for rapid intraoperative diagnosis of meningiomas. So frozen section is the current standard for intraoperative pathological diagnosis in neurosurgery, but it remains a labor and time-intensive process. Stimulated remodestology, or SRH, is a novel technology that may offer a streamlined approach for rapid intraoperative histopathological diagnosis, eliminating many of the steps associated with frozen section. SRH uses fresh, unlabeled, and unprocessed tissue in a simple one-step squash prep to assess the intrinsic vibrational spectroscopic properties of biological molecules in tumor specimens, such as lipids, proteins, and DNA, producing a digital frozen section type image. Previously, we have prospectively evaluated the diagnostic accuracy and time savings of SRH versus traditional frozen section in a general heterogeneous cohort of multiple tumor types. We found a statistically significant reduction across all tumor types in intraoperative diagnostic time of more than 30 minutes with SRH compared to frozen section. In addition to time savings, we found that SRH confers similar accuracy to frozen section across all tumor types. There was a 91.5% similar diagnostic correlation between SRH and the gold standard of permanent section, which was equivalent with a 91.5% correlation between frozen section and permanent section. And additionally, we also collaborated with researchers at the University of Michigan and Columbia University to create a convolutional neural network, which we trained on over 2.5 million SRH images. This convolutional neural network predicts brain tumor diagnosis in the operating room in under 150 seconds, an order of magnitude faster than conventional techniques. In a multi-center prospective clinical trial with 278 tumor specimens, we demonstrated that convolutional neural network-based diagnosis of SRH images was non-inferior to pathologist-based interpretation of conventional histological images with an accuracy of 94.6% for the convolutional neural network compared with the 93.9% accuracy for frozen section in human diagnosis. In the current study, we sought to focus our analysis of time savings and diagnostic accuracy for meningiomas specifically rather than a heterogeneous group of all CNS tumors with SRH compared to frozen section. In addition, we sought to provide a qualitative analysis evaluating the utility of SRH-mediated neuropathological console in differentiating between various meningioma subtypes as well as tumor grades. Following IRB approval, 82 patients undergoing resection for CNS tumors were enrolled in a blinded prospective cohort study. Of these, 26 patients were diagnosed with meningioma on frozen, SRH, or permanent section diagnosis and were included in the analysis. So our study design was split into two arms. We took a tissue specimen from each tumor and then we split it into two. For the SRH arm, we reported the elapsed time during SRH sample preparation as well as by interpretation by the blinded study pathologist and then we combined these two elapsed times. For the frozen section arm, the elapsed time during frozen section preparation and diagnosis, which we measured as from the time that the sample was sent off from the operating room to the time that pathology called the operating room with the diagnosis, was reported. Additionally, for both groups, diagnostic accuracy as well as determined by agreement of SRH and frozen section diagnosis with the gold standard of permanent section were recorded and compared. So interestingly, we found that SRH-mediated mean time to diagnosis was statistically significantly shorter than time to diagnosis with frozen section. On average, SRH took 26.6 minutes less than frozen section to arrive at a diagnosis. Additionally, agreement for SRH versus frozen histopathology and SRH versus permanent pathology were equal. The 95% confidence intervals indicated no significant difference in diagnostic agreement between modalities. McNamara tests showed no significant difference in ratings between modalities. So here we see classic pathological features of meningiomas that were clearly visualized on SRH, including meningothelial roles, syncytial cells, intranuclear pseudo-inclusions, and somatobodies. Further, the pathologist was able to distinguish between various meningioma subtypes with SRH, including the meningothelial variants, demonstrating classical roles in syncytial cells, the fibroblastic variants, demonstrating spindle cells and thick bundles of collagen, and the transitional variant, demonstrating both fibroblastic and meningothelial features. Here we see that SRH is able to demonstrate classical characteristic features of WHO grade 2 atypical meningioma. Panels A to C show mitotic figures. Panel D shows a hypercellular sheet-like growth. And panel E demonstrates prominent nucleoli, all of which are suggestive of grade 2 meningioma. So why is rapid and intraoperative pathological diagnosis of meningiomas important? Accurate diagnosis of meningiomas is critical for school-based cases where rapid margin identification and local structure invasion is important. As an example, we present a case of a 75-year-old male who underwent a left frontotemporal craniotomy in 2014 for the resection of a WHO grade 2 sphenoorbital meningioma. He presented four years later with new left-sided vision loss, proptosis, diplopia, and temporal bossing. MRI demonstrated a recurrence of the lesion with invasion into the temporalis muscle and the superior lateral orbital wall. During surgery, a large pterianal incision was made and a superior lateral orbitotomy was performed. The tissue was identified intraorbitally back to the orbital apex and superior orbital fissure. The tumor consistency was very soft, which made it challenging to distinguish the tumor from the orbital fat and lacrimal gland. Furthermore, the tumor was very adherent to the temporalis muscle, making dissection difficult without removing a significant portion of the muscle. As a result, we sent multiple samples that were taken from the margin of the tumor within the temporalis muscle, the temporal bone, and the orbital apex. We then analyzed them with SRH in an effort to prevent disruption of the lacrimal glands, superior and lateral rectus muscles, and the temporalis muscle. SRH, as you can see, is clearly able to demonstrate the presence or absence of tumor cells in various tissue types. In the first row, we see dark streaks showing large acellular collagen fibers representing bone fragments with active tumor cells around it. In the second row, we see lacrimal gland tissue with organized glandular tissue and lipid deposits with no tumor cells seen. And then finally, in the third row, large acellular pink areas represent lipid deposits within the temporalis muscle, again with clear tumor cells surrounding them. It's important to note that tissues don't always look comparable to typical H-knee staining. As we can see, bone and skeletal muscle look similar in SRH, but the tumor cells are very easy to identify, which is what's important for determining margin status. So the resection was then continued extradurally via dolan-kaku approach to expose the cavernous sinus and middle fossa floor, and the remaining tumor was removed from the middle fossa skull base. Afterwards, no active tumor was visualized in the orbital temporalis muscle on either SRH or visually by the surgical microscope. A small amount of residual tumor was intentionally left within the cavernous sinus in an attempt to prevent injury to any of its contents, and the surgery was uneventful and the patient had no complications with notably improved vision and diplopia three days postoperatively. So in the future, we aim to leverage our preliminary data, which we described here, to justify and perform a prospective pilot clinical trial in which we will establish a site-specific specimen intraoperative database for meningiomas and other tumors of the anterior skull base and sinonasal cavity. We will store the stereotactically determined three-dimensional location of each site from which tumor specimen was collected, along with SRH image information. We aim to use machine learning techniques to identify features and SRH images associated with prognosis and recurrence of anterior skull base and sinonasal cavity tumors. And with that, I'd like to thank all of my collaborators as well as my mentors, Dr. Michael Ivan and Dr. Jacques Morcos, for their help with this project. Thank you. Hello. Good afternoon. My name is Victoria Clark, and I'm a sixth-year neurosurgery resident at Mass General Hospital. I'm going to talk to you today about my research characterizing the RNA polymerase 2 meningioma driver mutations, how they dysregulate transcriptional pause release, and exhibit susceptibility to TDK9 inhibition. Three key points. Thematic mutations in the DOC domain of the catalytic subunit of RNA Pol II, which is encoded by the gene Polr2a, drive the formation of surgically challenging skull base meningiomas. That these mutations lead to dysregulation of promoter proximal pause release, a critical step in transcriptional regulation, and that transcriptional output in Pol II mutant cells is susceptible to TDK9 inhibition. So during my graduate school time, I focused on the genomic characterization of meningiomas, and we found that Pol II mutant meningiomas are enriched along the skull base with the higher amount of enrichment, especially at tuberculin cell up. So if you can harken back to your molecular biology days, you may recall that RNA polymerase 2 is arguably the most important protein complex essential for life. It's the complex that transcribes our genetic information encoded in DNA to messenger RNA, which can then go on to be translated. While Pol II is a multi-subunit complex, the catalytic subunit is a protein called RPB1, which is split into multiple domains, and in fact, this is the first time that RPB1 has ever been found to be mutated in human disease. The meningioma driver mutations are localized to a specific domain called the DoC domain. The function of the DoC domain is not completely clear, although we do know that it splits the catalytic active sites into two and is immediately adjacent to the RNA as it exits out of actively transcribing RNA Pol II. Since I began working on this problem during graduate school, there have been massive advancements in the structural understanding of transcription and especially advancements in the understanding of a process called promoter-proxy pause release. So RNA polymerase 2, after it initiates at the promoter, it does not switch immediately into active elongation. Instead, after transcribing for around 50 base pairs, it pauses, and this pause is regulated in part by two pause release factors, NELF and DSIF. With CDK9 phosphorylation, both of these factors and of the tail of RNA polymerase 2, now RNA polymerase is able to go into active elongation. With the emergence of these crystal structures of Pol II in complex with the pause release factors, I realized that the meningioma mutations in the DOC domain are immediately adjacent to part of the pause release factor DSIF, which is known to be critical for promoter-proxy pausing and is directly phosphorylated by CDK9 prior to PTFE, leading to my hypothesis that DOC domain mutations are impacting promoter-proxy pause release via alterations in this area of DSIF known as the RNA clamp. To assess this, I modeled the Pol II DOC domain mutations, which is the most common recurrent mutation in the mouse embryonic stem cell system using CRISPR-Cas9, and then I performed ChIP-seq for total Pol II as well as the phosphorylated versions and calculated what's known as a traveling ratio. So, the traveling ratio speaks to quantify what proportion of RNA Pol II is sitting at or near the promoter, that is, that which is paused, over that which is found along the gene body, which is elongating. So, if a given gene has a higher proportion of elongating polymerase, then in this ratio, the denominator would be higher, the numerator would be lower, and the traveling ratio for that gene would be a smaller number. If you take the total distribution of traveling ratios amongst all genes, if in fact the mutations and RNA polymerase two are impacting pause release, you would expect for the whole distribution of the traveling ratio to be left shifted. And in fact, that is what I found. So, by defining active genes either by ChIP-seq markers or just by looking at the top highest transcribed genes by RNA-seq, I do find that the Pol II mutant cells have a left shift in the traveling ratio reflecting a defect in transcriptional pause release. I next wanted to see if this defect or this phenotype could be rescued by CDK9 inhibition. But first, I needed to create a reporter cell line. I used the Tetra system with piggyback transposase in order to insert over 3.5 million copies of luciferase per cell line. So, these are stable cell lines. In order to assay, basically baseline transcription, the first thing I noticed is that the Pol II missense cells at baseline have about a 1.3 fold higher level of baseline transcription compared to the wild-type cells, consistent with the mutations promoting a pro-elongation state. Next, I screened several CDK9 inhibitors, including DRB, a non-selective CDK9 inhibitor, and AZD4573, which is selective and currently in phase I clinical trials for the treatment of hematologic malignancies. And I found not only is the selective CDK9 inhibitor effective, but it seems to have a higher impact on nascent transcription in the mutant cells compared to the wild-type, with a 40 or 50 percent decrease in transcription for the mutants, while only around a 27 percent decrease for the wild-type. I then utilized a new technique called FlamSeq, which is metabolic labeling of nascent transcription for a certain time point, followed by next-generation sequencing. This technique, in which 4-thiouridine is incorporated only during the pulse, allows one to assess not only the fraction of transcripts which are being nascently transcribed, but also the study states. And I found, using FlamSeq, that the Pol II mutant cells did have a significantly higher decrease in TC conversion, which is reflective of a significantly increased or significantly decreased amount of nascent transcription after treatment with a targeted selective CDK9 inhibitor compared to wild-type, with a 75 percent decrease compared to 67, although this was only looking at a single time point and a single concentration of AZD4573. The reality is, on the transcriptional level, there is a significantly decreased amount of transcription also in the wild-type phase. So, while this does show promise, ultimately, the dose optimization is ongoing. So, the key points, the somatic mutations in the doc domain of the catalytic subunit of Pol II drive surgically challenging skull-based meningiomas, that this leads to dysregulation of the process, a promoter of proximal pause release, and that the output is sensitive to selective CDK9 inhibition. I would like to thank, first, my research mentor, Rick Young at the Whitehead, as well as multiple collaborators, including Dr. Ganell at the Yale, MGH Neurosurgery, and especially the NREF, who funded my research on this topic. Thank you. Good afternoon. My name is David Bakshinyan. I have recently completed a PhD degree out of Dr. Sheila Singh's laboratory and currently pursuing an MD degree at the University of Toronto. First, I would like to thank the organizers for giving me the opportunity to present my work, as well as Synaptic Medical for selecting the abstract for the award. Today, I'll be speaking on our efforts to rationally develop synergistic therapies alongside BMI1 inhibitors for group 3 medulloblastoma. Without going into too much background on medulloblastoma, I just want to point out that although the multimodal therapies against medulloblastoma have rapidly evolved and advanced over the past years, it still remains one of the leading causes of pediatric mortality and morbidity. In the past decade, what was long thought as a single disease have been since stratified into unique molecular subgroups, each with a unique demographic and prognostic features. Out of the four key molecular subgroups, group 3s represent the most aggressive subgroup with worst clinical outcomes, as well as highest frequency of recurrence and thus the focus of our research. The aggressive nature of these tumors have been long attributed to this persisting population of therapy-evading cancer stem cells, or what we term them as brain tumor initiating cells. In our own research, we do see that when we look at BMI1 signature high tumors, the patients with those tumors tend to do much worse in terms of overall survival, as well as their propensity to recur after treatment. On a molecular level, what this presents is a treatment opportunity whereby we can go after those cancer stem cells that are evading therapy with precision small molecule targeting modalities and potentially eradicate those tumors or at least alleviate some of the long-term negative events. In 2019, we published a paper that has shown that indeed such small molecule targeting BMI1 does indeed reduce the proliferation and self-renewal capacity of the key properties of those cancer stem cells in group 3 medulloblastoma lines. In this case, we're using SUMB002, a line that was derived from a three-year-old male patient that has been treated only with cyclophosphamide. And those in vitro findings do translate to in vivo work as well, as we do see a reduced tumor burden in the treated mice. However, regardless of the dosage, mice still did succumb to their tumor burden. And that prompted us to ask a question, what would happen if we combine this BMI-1 inhibitor with already established standard of care. That's what exactly what we've done. We combined this targeted approach with a mouse-adapted therapy model of a patient-derived xenograft, whereby we treat mice with two gray of cranial spinal radiation, followed by a single doses of cisplatin increased in cyclophosphamide, the gold standard treatments for group three medulloblastoma. And in this experiment, we did divide mice into several arms and the three notable ones are, we combined standard of care alongside BMI-1 inhibitor in the concurrent matter, whereby we treated mice with BMI-1 inhibitor throughout chemoradiotherapy, or in a sequential fashion, whereby we completed standard of care treatment and then went in with a BMI-1 inhibitor. And in the panel B, the survival curves do show that combined treatment regimen did increase the overall survival compared to either of the modalities alone. It still wasn't sufficient enough to eradicate the tumor entirely. And that prompted us to ask a question, well, what's driving this therapy evasion? And we leveraged the established model to really employ two new techniques. One, cellular DNA barcoding, whereby we wanted to answer a question, what are the clonal dynamics? What is the heterogeneity status of the remaining tumors? And then second was genome-wide CRISPR screen to identify sensitizes and resistors to this BMI-1 inhibition alone and in combination with chemoradiotherapy. So that's exactly what we've done. We transduce the cells with a unique barcode library that comprised of 1 million unique barcodes. We transduce the cells, expanded the cells in vitro, and then xenografted them into immunocompromised mice, and then undertook the similar treatment regimens as I described before. What was interesting to observe right away was that there was about 10% engraftment rate as we were only able to recover about 10,000 barcodes from the brains and about 3,500 unique barcodes from the spine. So that does show that there's a massive bottleneck event occurring and not all cells are able to engraft and initiate tumor progression. But what was striking to see is that when we combined the treatment standard of care plus BMI-1, we were able to increase the frequency of unique barcodes that were the causing of this therapy evasion. And we do see a massive loss of tumor heterogeneity, suggesting that a polyclonal, suggesting a polymultimodal therapy approach might be sufficient to eradicate those tumors. To address the question of, well, what can we combine with BMI-1 inhibitor? We undertook an in vitro genome-wide CRISPR Cas9 screen and we identified the sensitizers and resistors to this BMI-1 treatment. And when we looked at the pathways or different protein networks that are up-regulated or down-regulated after a BMI-1 inhibition, we came up with four key regulatory pathways that kept coming up, PI3K, BLK1, mTOR and Aurora kinase B. So in the current set of experiments, what we've done is while we've profiled all those, we selected small molecule inhibitors to go alongside BMI-1. And I'll direct your attention to Ensozurin, a PI3K inhibitor. And what we see with that, when we test its IC50 or inhibitor concentration of 50, sufficient enough to kill 50% of the cells, we do see a sensitization effect whereby treating naive MB002s show the value of 23 micromolar, whereas the MB002 cells that were treated with thunderscare and radiation, their IC50 is about tenfold lower, suggesting that those cells are primed for that inhibition. And in terms of BLK1 inhibitor, that was another class of inhibitors that showed promise to be combined with BMI-1 inhibition. And we see a large therapeutic window with both PI3K and BLK1 inhibitors. And in case of medulloblastomas that were collected at endpoint after standard of care or chemoradiotherapy alone, we see almost a 66,000 fold difference in the IC50 values. And our early experiments testing this synergistic effect definitely did show that in combination, BMI-1 and PI3K inhibitors do synergize and lead to a greater reduction in cellular proliferation as read out by cell viability assay. As you can see, we've tested three different patient-derived lines, MET411, MB002s, the cells that were originally used in our screen, and D425, a commercially available cell line extensively validated group three cell line. So in summary, BMI-1 alongside standard of care does extend survival in mice, but it is insufficient to eradicate tumor progression and mice do succumb to their disease. The combinatorial treatment regimen do reduce clonal heterogeneity, and PI3K pathway represents a rational target to combine with BMI-1 inhibition. And our current experiments are focused on expanding this in vitro synergy studies to additional group three lines, as well as group four medulloblastomas and contrast those results with human neural stem cells, as well as initiating of in vivo studies using PTC596 or BMI-1 inhibitor alone and in combination with those inhibitors. With that, I would like to thank everybody for their attention, my supervisor, Dr. Sheila Singh, all the patients and their families for their generosity, and of course, so my committee members and mentors. Thank you very much. Thank you. Great. I wanna thank the first three speakers. We're having a little bit of technical difficulty with Dr. Rinda's presentation, so we're gonna move on to Dr. Pomerantz, so we can proceed with his presentation. If we can get things sorted out, Dr. Rinda will be the last presentation of the day after Dr. Burns. Thank you. Okay, great. Thank you for the opportunity to talk today. I'm a resident in the joint UVA-NIH program, and the presentation today will cover the effects of targeted radiation to the hypothalamic pituitary axis and the effects of radiosurgery on pituitary tumor control and endocrine function. We'll start with a few key takeaways. We performed a multicenter study from treatment centers across the world as part of the International Radiosurgery Research Foundation and this is actually a follow-up study to an original investigation performed at UVA. And our goal was to evaluate the radiation tolerance of structures surrounding pituitary adenomas and identify predictors of delayed hypopituitarism following radiosurgery for these tumors. Over a five-year follow-up period, tumor control was achieved in 94% of patients. We found that doses to the pituitary stalk with a threshold of 10.7 gray, as well as doses to the normal gland significantly increased the risk of delayed hypopituitarism. However, not all endocrine axes were equally affected in terms of delayed treatment effects. And there's still an unmet clinical opportunity to define adaptive dosimetry parameters that are patient and pathology specific. And in order to meet this need, this study represented the largest international cohort with multivariate dosimetry analysis and endocrine function after radiosurgery. We know that radiosurgery works well for controlling residual or progressive disease, really in patients that can't be cured with surgery alone. And in general, tumor control is around 90%. And we also know that new or delayed endocrinopathy occurs in about 20 to 30% of these patients. And while the idea of dynamic radiance of radiation has been explored, we really don't know what specifically leads to delayed endocrine deficits based on targeted radiation to different structures. So we performed a review of patients with pituitary adenomas who underwent single fraction stereotactic radiosurgery at 16 institutions. Point measurements of doses along 14 predefined structures of the hypothalamus, pituitary stalk and pituitary gland were made. And statistics were performed in order to determine the impact of doses to critical structures on clinical radiographic and endocrine outcomes. Pituitary function before and after radiosurgery was assessed through the serum markers. And for this study, hypopituitarism was defined as deficiency of hormones below normal limits and patients who required medical replacement. Clinical and radiographic information from the day of radiosurgery served as index data for relative changes during follow-up. This is actually an illustration of the 14 points along the HP axis. These points were selected on stereotactic planning MRI and analyzed using gamma plan software. And as you can see, there are three points along the left and right hypothalami, five points along the pituitary stalk and three along the pituitary gland. And so these descriptions were provided to all participating centers so that data could be obtained consistently. And point dose measurements remained blinded to outcome, including tumor control and endocrine outcome. Overall, 521 patients were included in the final analysis. There were equal, approximately males and females. Median follow-up was just over five years and most patients underwent prior transphenoidal surgery and most tumors were non-functioning. Treatment parameters across institutions were largely consistent. The relatively low dose to the hypothalamus really follows prior recommendation to limit radiation there. Some groups have argued that radio sensitivity may vary among hypothalamic nuclei. And so certain hormonal axes may become deficient before others. The doses to ventral structures were higher with median doses of 7.2 gray to the pituitary stalk and 11.3 to the normal pituitary gland. Before radiosurgery, about 70% of patients had hormone insufficiency. After radiosurgery, 24% of patients developed new hypopituitarism and this was due either to treatment effect or tumor progression. And there was a high rate of tumor control, again, in 94% of patients. The number and types of endocrinopathy after radiosurgery were classified by endocrine axis and of 124 patients with new endocrinopathy, you can see a steep fall off that occurs from patients who had just one affected hormone axis to those with two or more. And this implies that a treatment gradient existed with respect to anatomical terrain and possibly hormone physiology. Factors associated with new pituitary insufficiency were identified using logistic regression. The takeaway here is that non-functioning adenoma, younger age, higher margin dose and higher dose to the normal gland as well as the stock were all significantly associated with delayed endocrine dysfunction. We found that the maximum dose to the normal gland as well as the stock increased the likelihood of long-term endocrinopathy. This finding highlights the principle of shielding healthy vital structures from differential threshold irradiation and our results support the concept of the pituitary gland in the stock as discrete anatomic entities. And we found that a median dose threshold of 10.7 gray delivered to the stock actually protects against hormone deficits. So newer worsening endocrinopathy was further analyzed by specific endocrine axis in a multivariate logistic regression. The takeaway here is that not all axes were equally affected in terms of delayed treatment effects. Higher dose to the normal pituitary gland was a predictor of dysfunction of every endocrine axis except for prolactin. And on the other hand, maximum dose to the normal gland was really only predictive of thyroid deficiency. There were several key limitations of this study. First and foremost, there was systematic bias due to differences of the treating institutions, including selection bias and variability of time between diagnosis and treatment. Subjectivity was also introduced likely when determining precise treatment volume as well as the actual point doses, especially for patients with unclear tumor margins and other structured boundaries. And finally, hypopituitarism was defined relative to baseline endocrine status prior to radiosurgery. And since the majority of patients had deficits before treatment, not all endocrine axes were susceptible to the same degree of risk for each patient. So we need to study dose thresholds and dose ratios in patients with versus without preexisting endocrine deficits, as well as with and without prior radiation, which may both be confounding effects in terms of understanding the drivers of delayed endocrinopathy. In conclusion, radiosurgery for pituitary adenomas offers a high rate of tumor control and a relatively low risk of new or worsening endocrinopathy. Our evaluation of point dosimetry to adjacent structures revealed that doses of the pituitary stock with a threshold of 10.7 gray, as well as doses to the normal gland significantly increased the risk of hormone deficits. But our analysis also suggested that various endocrine axes have different sensitivities to point doses of radiation. So using appropriate dose planning, it may actually be feasible to refine stereotactic radiosurgery delivery, not only to accomplish tumor control and endocrine remission, but also offering a better chance of preserving normal endocrine function. We had tremendous help from a broad list of collaborators across the world. I'd like to thank them, the International Radiosurgery Research Foundation and my mentor at UVA, Jason Sheehan. Thank you. Hello everyone. My name is Irakli Abramov and I'm a research fellow at Barrow Neurological Institute. And today I would like to share with you our experience in using confocal laser endomicroscopy for intraoperative discrimination of tumor histoarchitecture. No disclosures. Confocal laser endomicroscopy or CLE is an emerging technology in neurosurgery that allows for in vivo cellular discrimination of tumors. And we, in this study, analyzed the first FDA approved new generation CLE for intraoperative in vivo use during neurosurgical operations. Briefly, CLE system consists of a handheld miniaturized optical laser scanner that is designed as a rigid probe with a four millimeter outer diameter and a working distance of 150 millimeter lengths. This probe is connected to the operative control workstation where images are displayed. Image generation process is dependent on the intravenously injected sodium fluorescein which accumulates in the regions of blood-brain barrier disruption. The laser in the probe provides incident excitation light and the scanner mechanism detects a fluorescein emission which is conveyed via optical fiber to the detector. The detector signal is digitized synchronously with a scanning to provide images of tissues parallel to the tissue surface. CLE imaging was used among four neurosurgeons that were involved in the study. Different images were obtained from different regions of interest with a neuro-navigation confirming location of the probe. In 11 cases, CLE was employed in tandem with a recently developed telepathology software that allows in real time intraoperative consultation with a neuropathologist. Images were classified as interpretable and non-interpretable. Interpretable images were reviewed by an experienced neuropathologist that compared descriptively the CLE images to the corresponding HNE sections. Sensitivities and specificities were calculated for frozen and CLE images with HNE sections as a standard. A total of 30 patients were enrolled in our study and final diagnosis was consistent with primary brain tumors in 23 cases of which 13 were gliomas, secondary brain tumors in four cases and reactive gliotic brain changes in four cases. Next, I would like to briefly show you some of the examples of in vivo CLE application and this is a male, a 23 year old male that had a tactile glioma. A CLE imaging was able to detect hypercellularity with atypical cells that matched to the HNE sections and the permanent diagnosis for this patient appeared to be pylocytic astrocytoma. Another case was a 43 year old female that had a left sphenoid wing meningioma. CLE was able to detect so-called nest of tumor cells located at the border of a normal dura and this appeared to be meningioma grade two. This was an interesting case in a 58 year old male that had a left parietal lobe lesion that was regarded to be a recurrent glioma. However, CLE imaging was able to detect predominantly hypercellular regions that were consistent with microphages and you can see them demarcated with a curved white line. Demarcated with a curved white line. Another sign of reactive brain tissue that CLE solution was able to detect is so-called corporal amylation that you can see marked with white arrowheads and that is generally a sign of neurodegeneration changes. This was a particularly interesting case of a 28 year old male that had a non-enhancing diffuse left frontal lobe glioma that one would expect to be a low grade glioma. However, CLE imaging was able to detect a vast majority of hypercellular cells and ATPO cells that was consistent to be with an anaplastic astrocytoma. Every technology has its own advantages and CLE imaging is not an exception. Of total images acquired, 54% of images were non-interpretable and this fact was related mainly to the motion artifacts and red blood cells obscuring the field of view. However, we observed a strong positive correlation between percentage of interpretable images and timing of CLE system usage in the OR. That would definitely mean that there is a learning curve and especially assuming that majority of surgical team members were not experienced CLE users. We found that superficially located lesions appear to be less interpretable than deep located lesions. At the same time, meaning gliomas had a lower percentage of interpretable images. CLE showed lower sensitivity when compared to frontal sections with tumor tissue discrimination. However, there was observed a higher specificity for reactive brain tissue changes. Surveys that was conducted among neurosurgeons involved in the study showed that the main disadvantage of system was adherent with the motion artifacts and red blood cells obscuring the field of view. However, all responders were consistent in ranking the remote pathology platform as a very useful guidance tool during surgical operations. In conclusion, I would like to say that this study showed that a new generation CLE system allows in vivo pathological tissue discrimination. However, red blood cells and motion artifacts can significantly impair tissue discrimination. And I would like to thank every person who helped throughout the study, including research fellows, residents, and especially my mentor, Dr. Mark Proulx. Thank you. Good afternoon, everyone. Thank you for the opportunity to speak and appreciate the program organizers and the award selection committee choosing this talk for the American Brain Tumor Association. So we've been very interested in this problem, which is that brain tumors keep coming back, particularly gliomas. And so we know that the decisions that are made by tumors can be traced to biochemical processes. And those can range from decisions about proliferation, decisions about dying, decisions about how to escape therapies. And so each of these biochemical phenomena have a fingerprint. And you can see here a picture of a cell. Obviously there's a lot of complicated things going on inside the cell, but outside the cell, there is also a reflection of what's going on inside the cell. And so if you sort of take a step back and think about why are we failing so miserably? I think part of the problem is we don't get any feedback when we give a therapy. All we have are clinical trials where patients are randomized. And at the end of the day, you have a survival time and it's either shorter or it's not shorter without treatment, but you don't have a mechanistic answer of how can we improve. And so that's what we want to be able to accomplish, to get feedback. And so in preclinical models, we can place catheters in brain tumors and we can measure things and we can find that there are biomarkers that correlate with certain types of tumors. But preclinical data don't have a great track record of predicting what's going to work in patients. And so as a step towards that goal of getting better feedback longitudinally from the human brain, we thought, let's give it a shot. Let's see, can we get some of this fingerprint metabolomic data back from the human brain in a live human brain tumor? Because we know that what typically happens is you take the tumor out and it goes in a bucket and you pour fixative on it. And you don't then have an opportunity to interact with it, but we wanted to change that. And so this is a snapshot intraoperatively of a pilot study we're doing where we're placing microdialysis catheters into the live human brain tumor while we're doing mapping prior to the resection. And so what is microdialysis? It's a strategy whereby you've got a pump and you can see the three pumps here wrapped in sterile bags and those are perfusing fluid through these little tubes. And there's a small catheter at the end and the location here you can see, we've got one enhancing tumor, one non-enhancing tumor and one in brain adjacent to tumor. And there's a semi-permeable membrane that allows solutes and metabolites to diffuse in and out of that membrane. And so at the end, you get an equilibration and a reflection of what's actually going on. And then as the pump keeps pushing the fluid back, it ends up in this vial and we can collect it. And so we had five patients in a first pilot study, a couple of oligodendrogliomas and some glioblastomas. We only had one patient here that we knew the diagnosis ahead of time because the patient had had a prior IDH mutant tumor and it looked bad, so we presumed it was gonna be a glioblastoma now. The others, we didn't know ahead of time what we were going to find. And so we know that IDH mutant tumors create a huge amount of 2-hydroxyglutarate. And so we started with that and we asked, does the IDH mutant recurrent glioblastoma and these oligodendrogliomas have a higher level of 2-hydroxyglutarate than these IDH wild type glioblastomas? And the answer is, well, yes, they do. In fact, if you look at the log fold change, it's actually over 10 fold higher. And interestingly, there was this one patient here whose pictures were on the original screen here. And it's actually almost a thousand fold higher than baseline we're finding in the non-enhancing flare positive tumor. And so there are high levels of metabolites that can be measured in the extracellular fluid and this is done essentially within 40 minutes during an operation. And so then we asked, what about the rest of the metabolites? And we found that there are 181 metabolites and we did principal component analysis and we found, oh, to our surprise and dismay, there was absolutely no difference based on looking at the whole global signature between IDH mutant and IDH wild type tumors. And so we thought, oh, well, goodness, maybe we're trying to do things too quickly in the operating room, usually for people doing interoperative microdialysis, you let things equilibrate over days and you're measuring drug levels. And so we went back to a simpler question and asked, what about the enhancing tumor versus the non-enhancing tumor? And we made a ranked list of the metabolites that are most enriched in the enhancing tumor versus the non-enhancing tumor. And we found that we could identify that signature in other patients, suggesting that with a single patient, we could find metabolites that are enriched in enhancing tumor and predict what was going to be enriched in another paradigm. But these are very diverse tumors. And so, although we've got a Pearson correlation showing that there's this interaction between the metabolites, we found that each tumor is different. And so the patient samples actually did cluster by patient. And if you look at the PLSDA plot, there seemed to be a breakdown between a couple of tumors here, which couldn't be separated and others here. And it was interesting because there were two oligos, but then also this IDH wild type GBM that was separate from these other two tumors, one of which was IDH mutant and one of which was not. And this was the principal component analysis before the PLSDA. So again, if we break things down by IDH mutant versus IDH wild type, it's not very obvious that there's a huge difference. But if we break it down by this division that we observed, these two groups, and we described the one with the single IDH mutant tumor and the one IDH wild type gliostoma as alpha, and the others as beta, there was a marked difference between them based on the empiric observation. And so if we do the volcano plot, we see that we're not surprised to see two hydroxyglutarate there. There's not very much else up there. And if you look at the log fold changes, we're sort of maxing out at about four. But if we look at this other phenotypic difference, we see that it's up to a log fold change of eight. If we compare each of those to brain, we found that there was some metabolites that overlap, suggesting that each of these tumor phenotypes do have a common signature that we can identify, but then there's also a unique characteristic. And although we need more patients to define it, what it appears is that actually the beta phenotype is very, very similar to brain itself, but much more highly upregulated, as though the tumor is trying to do exactly what the brain is doing, but more of it. But this other phenotype, including the recurrent tumor, looks different, and it's probably reflective, although we need to confirm this with additional patients of the Warburg switch from oxidative to glycolytic phosphorylation, which we can measure now within 20 minutes to 40 minutes per sample in the operating room. And that gives us an opportunity. We're measuring other drugs here, the Keppra, the Kefzol, we see a correlation between the enhancing tumor and the levels. And I envision a future where we could actually be delivering candidate drugs through catheters that are placed within the tumor, and seeing how does metabolism change in response to that? And so we did a pilot study in animals. We gave an IDH inhibitor. Within nine hours, we could completely wipe out the 2-hydroxyglutarate. And so I think there's opportunity here to come up with new paradigms for therapies. We can potentially come up with implantable strategies, and we can also monitor CSF as well for longitudinal strategies. So I'll leave it there. Lots of thanks to people who are part of a big team, but I just want to leave you with this quote that's outside of where I see patients during, in between surgeries in the outpatient clinic at St. Mary's Hospital. We must not be content to see things as they are. We must have the vision, faith, and hope to see things as they can and must become. And so I leave that with you to see what else can we learn during the surgeries where we have access to this tissue that we're so privileged. And we thank our patients for that and their various funding sources, including the American Brain Tumor Association. Thanks so much. Thanks very much, Terry. Thanks to all the speakers in this afternoon session. Unfortunately, we couldn't get things sorted out with Dr. Rinda, so we won't be able to have his presentation today. But once again, all the talks were just excellent. Thank you everyone for tuning in to this virtual session today. As I'm sure you all are, I'm very sad that we can't be together in person, but I just have to applaud the AANS leadership for putting this together and all the amazing people behind the scenes who are making this happen so we can still at least virtually be together and learn. And hopefully, hopefully we'll all be together at the next AANS in Philadelphia in April 2022. So thank you once again, have a good evening, and enjoy tomorrow's program as well.

endoscopic approaches

skull base

cavernous sinus

middle wall resection

anatomy of the skull base

tumor removal

surgical approaches

complete tumor resection

precise anatomical knowledge

posterior petrosectomy approach

wide exposure

cranial nerves

patient safety

microdialysis

innovative technologies