Transcript
All right. Well, thank you very much. Thank you to Brainlab and Novalis Circle organizers for having me here and giving me the opportunity to speak with you today. This talk is gonna be a little bit different than all the other talks we've heard so far. I know this is a radiosurgical conference. I'm a neurosurgeon who performs operations and radiosurgery for these tumors. And we do treat at my institution, many skull-based tumors, meningioma, schwannomas with radiosurgery.
But there is always an instance where patients need surgery because of large enough tumors that have sufficient mass effect or associated symptoms and decompression is needed. And for that subset of patients, I wanna talk about what tools we have in the operating room to improve our visualization and make surgery better, and also then start the pre-planning process for adjuvant radiosurgery for residual that we're gonna leave behind. And a lot of that kind of interacts with the adaptive hybrid surgery model that you just saw.
You know, by way of disclosure. Unfortunately, I have nothing to disclose. I would love to own a piece of Brainlab, but Stephan has not asked me to invest. So, you know, when we talk about improved visualization, the reason that we have advanced imaging tools, intraoperative imaging navigation in the operating room is actually severalfold. It allows us to have better localization and better orientation when we're working in small windows, looking at complicated anatomy, but it also allows us to pre-plan and determine the approach we wanna take, particularly in this modern era of trying to make everything more minimally invasive. The more minimally invasive an approach you take, the more thoughtful you have to be about the way you approach it and exactly the windows and corridors that you have to work in.
Also, we wanna be able to assess our progress as we're operating. In this modern era, we understand that the goal of surgery for benign tumors is not necessarily gross total resection, but getting us to a point where symptoms are controlled and then using other tools such as radiosurgery to give us long term tumor control. But we have to know as we're operating how well we're doing, how much tumor have we removed, what residual exists and where is that residual?
And so we can use either intraoperative true imaging, such as intraoperative MRI, or CT, or we can use image-guided navigation tools to give us a rough estimate if we're not so lucky as to have those imaging tools in the operating room. We have to integrate all of that information together so that we can use this information to plan both the end of the operation and then the adjuvant treatment we're gonna give. And we wanna do it in an efficient way to minimize the time and exposure of the operation.
And so we have a number of tools that are today available and in operating rooms around the world. The most basic of which has been around for decades is basic image-guided or stereotactic navigation. The tools have been getting better as time goes on, but this is really technology that has been in most academic centers since the 90s and almost every center in the modern era.
Some people are lucky enough to have inoperative imaging. There are a handful of centers that have inoperative MRI, and that can be immensely useful, but also extremely expensive. A few more centers have inoperative CT, which is a little bit more cost-effective, but it's not technology that's always available to us. And so hopefully I'll show you that with stereotactic navigation, we can make up for what we lack in real-time imaging.
We now have new tools with virtual reality and augmented reality where we can import information from our advanced imaging, from our stereotactic navigation, and actually bring it into the field through our microscopy and endoscopy, and that can enhance our visualization and help us perform these operations. And then we have adjuvant treatment planning actually occurring in the operating room. And that's the adaptive hybrid surgery platform that you just heard of.
So basic frameless stereotactic navigation, for those of you who are not familiar with it takes tools similar to what you use in radiosurgical planning. So importation and fusion of imaging, CT, and MRI, identification of structures using Atlas-based segmentation and then manual painting of objects, such as the tumor or at-risk structures that we're concerned about. And then it combines that with cameras that can actually visualize the patient and the instrument.
And so after we've gone through a registration step, the system knows where the patient is in real time, and we can navigate with instruments in the patient's anatomy and see how that lines up with preoperative scans and identified imaging acting sort of as a GPS within the brain. And this is old technology that's helped us to determine where to make a craniotomy, how to approach a tumor, and has been used just in of itself to estimate extent of resection in the past. But we're getting better. We have better tools now.
We can combine that with intraoperative imaging, as I mentioned, CT or MRI, and we can take all of that information that we have, and we can place it into our field division as we operate using augmented reality. So now with the ability to identify three dimensional structures, the tumor, fiber tracks, organs at risk, we can import those as three dimensional objects through the microscope viewfinder and actually see superimposed objects that are difficult to define by vision alone, marked out in order to help us see these borders. And I'll show you that in actual cases.
And then lastly, there's this idea of intraoperative adjuvant treatment planning. So we can use intraoperative structure updates to estimate how much tumor is left behind and then allow the computer system to do a quick assessment of the quality of a radiosurgical plan to help guide the surgeon intraoperatively as to how much to resect and when enough has been resected when we can stop and avoid the morbidity of taking out that last 5%, 10%, 20% of tumor.
So I wanna give you examples of how we use this to tackle the different areas of inoperative planning and procedure that I mentioned. And I wanna start with the idea of minimally invasive surgery and approach selection. So here's a case of a patient of mine with a small to medium-sized plain sphenoid meningioma, but that was associated with significant frontal lobe edema who presented with seizures. And, you know, you could argue to treat this with radiosurgery, but because of the edema and because of the patient was symptomatic from recurrent seizures, we felt that it was appropriate to resect this.
So this patient, of course, like all patients wants a minimally invasive but excellent outcome surgery. And so how can we perform that? So there is a standard bifrontal craniotomy that we can use to approach this, but there are some better techniques that we have surgically today. One option would be to approach this endoscopically through a transnasal approach, which is actually an excellent way to get at that tumor. But this patient has intact olfaction and this approach almost always sacrifices olfaction, because you have to go transcribiform. And so there are also super territorial or, you know, approaches from above the skull base that can allow us to save the olfactory nerves.
We can approach through a minimally invasive supraorbital craniotomy with a small incision in the eyebrow and do a keyhole craniotomy or we can sort of extend that into a larger cranial orbital keyhole craniotomy that involves removing the orbital bar. But how do we know what the right approach is? Well, we planned a cranial approach because of her intact olfaction and we planned an eyebrow incision for this patient. And here you can see we're looking through the microscope at the patient. The proposed incision site is marked, and we can superimpose the tumor object from the navigation onto the patient to really have a sense of where that is in 3D and what the trajectory is to it.
And you really can't tell what you're gonna need until you get a little bit further into the surgery. But when we actually open that eyebrow and take a look here, we see that, you know, the simplest approach would be a supraorbital craniotomy, but really that would give this approach which doesn't get us down to the base of the tumor, doesn't allow you to get that dural base and achieve a real Simpson grade one or two resection. If we do a larger cranial craniotomy that really gets us down all the way to the base of the tumor.
And so you can see very quickly, it becomes apparent what approach we need, and we do this before we make the craniotomy so that we're not struggling afterwards. So this is, you know, what we achieve. We can remove the orbital rim and the frontal bone in one piece. And then if we look straight down our angle of view and navigate off the microscope, we see that the field of view off the microscope gives us a direct shot right to the base of the tumor.
And so that's gonna be sufficient to be able to take out that tumor. Indeed, we're able to achieve a gross total resection. With these minimally invasive approaches patients heal quickly and are able to return home very fast. This is the patient at a two-week postoperative visit, minimal pain, excellent recovery.
So that is how we use approach selection with intraoperative imaging. Now, what can we do with that added augmented reality that intraoperative visualization to improve postoperative adjuvant planning? And this is where we can combine that augmented reality with adaptive hybrid surgery. So this is an example of a patient that I just recently operated on last week with a petrous apex meningioma.
This is a non-contrasted image because this patient has chronic renal insufficiency and can't receive contrast for a contrasted MR which also tells us that if we leave any residual behind, we're gonna have some difficulty assessing the boundaries of that residual tumor. And if we do wanna treat with adjuvant radiation, we're probably gonna have to over target a little bit to make sure that we cover the area.
And if that's the case, then we need to know that we're very safe in terms of our residual away from at-risk structures, such as the brain stem, optic nerve, because this does go into the posterior cabinet sinus there. And so we can identify the objects, identify our plan ahead of time, and then we can go into the operator room. So this shows you the approach through a retrosigmoid craniotomy, and the line of site here that we get from the microscope right to the edge of the tumor. But what you actually see in reality is this.
So when you perform that retrosigmoid craniotomy, you see a small amount of tumor here sticking out between your trigeminal nerve and your vestibulocochlear nerve. And this is your working window. It seems like you have a good view of the tumor here, but if you actually use augmented reality to superimpose the entire tumor boundary over what you're seeing, you realize that there is a much more tumor than the window that you have to work through.
And this is very helpful. Even though we know this as surgeons, you know, there's no surprise here if you're experienced with this. It still helps to have a visual image of the actual boundaries, especially the upper and lower boundary that are sitting directly underneath those nerves. And so this helps us actually perform a more extensive resection to work piecemeal through this area until we've removed nearly all of that tumor. We're not gonna chase it forward into that posterior cave sinus. But once we've achieved what we think is safe and resectable, we can then use the intraoperative structure update that you saw to get an assessment of how much tumor has been removed. And we actually do this sequentially throughout the tumor resection. But I'm just gonna show you sort of the final resection here.
And ultimately, what we've achieved is a resection of everything down to the base of the, you know, petroclival angle and the extension going into the cavernous sinus. And we can superimpose our residual then back into our microscopic view and see what's left. And if we actually perform an adaptive hybrid surgery run, and we see that we're still too close, we can physically see where that area is that is still a problem for radiosurgery, so that we can target that with the operation.
So here's the adaptive hybrid surgery look. This is the initial tumor. Obviously, too big and too close to the brain stem for single fraction or even really hypofractionated radiosurgery initially. But after we've performed our resection, the residual looks like an excellent target, and we know we can stop here.
And then, you know, it was mentioned that the quality of the radiosurgical or the adaptive hybrid surgery information is dependent on the quality of the inoperative structure update that we get from the system. And so it's always important for us to assess how well we're doing here. This shows you the pre and postoperative images with the initial tumor and the residual tumor marked. And we see that the residual tumor actually matches up very nicely with the residual that's there on the postoperative image. And we look at that with every case to show that we're usually within about 5% volumetrically based on what the system suggests we have as a residual and what we actually get on the postoperative MR.
So how else can we use inoperative imaging and inoperative structure update to help us? So sometimes in the case of adaptive hybrid surgery, we want to intentionally leave some tumor behind and we wanna know can we radiate it. Other times, we wanna be very clear that we're removing all of the tumor. And I think that becomes very important in the special case of hyperostosis meningiomas of the skull base.
So the hyperostosis, this thickened bone can be very hard to differentiate its boundaries intraoperatively, and can also involve areas near cranial nerve foramen and make it very difficult for us to access. But it is important to remove this because often it is the hyperostosis in these cases that's causing cranial nerve compression and the patient's actual mass effect and symptoms. And also we know that this hyperostosis harbors live tumor and needs to be treated either surgically or with adjuvant radiosurgery in order to really get long-term tumor control.
So, you know, this is a case of a patient where you can see the hyperostosis in the sphenoid wing actually leads to orbital proptosis, which is both a cosmetic issue, but increases pressure on the optic nerve through the optic canal and within the orbit and leads the vision loss. And if you don't remove this completely, you just remove the soft tissue component here. This patient's vision is not gonna get better after surgery.
And so we can utilize the combination of preoperative MRI because I don't have intraoperative MRI at my institution, but we do have an intraoperative arrow CT, and we can fuse those images, CT and MR, to get a look at not only the soft tissue component of the tumor but also the bony component of the tumor, which is much better visualized on CT.
And then we can use our image painting tools to identify as separate objects, not only the soft tissue component of the tumor, but also the bony component of the tumor. And then combine those two to actually look at the relationships between the hypersonic bone and the soft tissue component. And we can use the intraoperative structure update then as we go through to look at how we're doing with our bony removal, completely separate from the soft tissue component of the tumor. So that see, we started with this large area of hyperostosis, and then we can use the inoperative structure update to go through the bone as we are resecting it, and estimate how much of that hyperostotic bone is left.
And again, when we look at the final outcome, we see that what is shown is the original tumor and the residual tumor matches up very nicely between the inoperative MRI based on the preoperative scan and the postoperative scan, which shows we've actually achieved the resection we wanted to. And so these are ways in which, again, inoperative imaging and these tools can help us maximize the goal of our surgery.
So in conclusion, you know, I think it's important that we utilize these intraoperative image guided navigation tools which can help us plan more minimally invasive procedures when appropriate, it can help us limit the extent of our resection so that we're being maximally safe but also maximally effective in terms of long term tumor control, and occasionally to maximize the extent of surgical resection when that is the goal of our surgery. And the integration of this information with the knowledge that we're gonna treat with adjuvant radiosurgery really helps us to perform optimal surgeries for this combined multimodal therapy. So I thank you very much for your time.
But there is always an instance where patients need surgery because of large enough tumors that have sufficient mass effect or associated symptoms and decompression is needed. And for that subset of patients, I wanna talk about what tools we have in the operating room to improve our visualization and make surgery better, and also then start the pre-planning process for adjuvant radiosurgery for residual that we're gonna leave behind. And a lot of that kind of interacts with the adaptive hybrid surgery model that you just saw.
You know, by way of disclosure. Unfortunately, I have nothing to disclose. I would love to own a piece of Brainlab, but Stephan has not asked me to invest. So, you know, when we talk about improved visualization, the reason that we have advanced imaging tools, intraoperative imaging navigation in the operating room is actually severalfold. It allows us to have better localization and better orientation when we're working in small windows, looking at complicated anatomy, but it also allows us to pre-plan and determine the approach we wanna take, particularly in this modern era of trying to make everything more minimally invasive. The more minimally invasive an approach you take, the more thoughtful you have to be about the way you approach it and exactly the windows and corridors that you have to work in.
Also, we wanna be able to assess our progress as we're operating. In this modern era, we understand that the goal of surgery for benign tumors is not necessarily gross total resection, but getting us to a point where symptoms are controlled and then using other tools such as radiosurgery to give us long term tumor control. But we have to know as we're operating how well we're doing, how much tumor have we removed, what residual exists and where is that residual?
And so we can use either intraoperative true imaging, such as intraoperative MRI, or CT, or we can use image-guided navigation tools to give us a rough estimate if we're not so lucky as to have those imaging tools in the operating room. We have to integrate all of that information together so that we can use this information to plan both the end of the operation and then the adjuvant treatment we're gonna give. And we wanna do it in an efficient way to minimize the time and exposure of the operation.
And so we have a number of tools that are today available and in operating rooms around the world. The most basic of which has been around for decades is basic image-guided or stereotactic navigation. The tools have been getting better as time goes on, but this is really technology that has been in most academic centers since the 90s and almost every center in the modern era.
Some people are lucky enough to have inoperative imaging. There are a handful of centers that have inoperative MRI, and that can be immensely useful, but also extremely expensive. A few more centers have inoperative CT, which is a little bit more cost-effective, but it's not technology that's always available to us. And so hopefully I'll show you that with stereotactic navigation, we can make up for what we lack in real-time imaging.
We now have new tools with virtual reality and augmented reality where we can import information from our advanced imaging, from our stereotactic navigation, and actually bring it into the field through our microscopy and endoscopy, and that can enhance our visualization and help us perform these operations. And then we have adjuvant treatment planning actually occurring in the operating room. And that's the adaptive hybrid surgery platform that you just heard of.
So basic frameless stereotactic navigation, for those of you who are not familiar with it takes tools similar to what you use in radiosurgical planning. So importation and fusion of imaging, CT, and MRI, identification of structures using Atlas-based segmentation and then manual painting of objects, such as the tumor or at-risk structures that we're concerned about. And then it combines that with cameras that can actually visualize the patient and the instrument.
And so after we've gone through a registration step, the system knows where the patient is in real time, and we can navigate with instruments in the patient's anatomy and see how that lines up with preoperative scans and identified imaging acting sort of as a GPS within the brain. And this is old technology that's helped us to determine where to make a craniotomy, how to approach a tumor, and has been used just in of itself to estimate extent of resection in the past. But we're getting better. We have better tools now.
We can combine that with intraoperative imaging, as I mentioned, CT or MRI, and we can take all of that information that we have, and we can place it into our field division as we operate using augmented reality. So now with the ability to identify three dimensional structures, the tumor, fiber tracks, organs at risk, we can import those as three dimensional objects through the microscope viewfinder and actually see superimposed objects that are difficult to define by vision alone, marked out in order to help us see these borders. And I'll show you that in actual cases.
And then lastly, there's this idea of intraoperative adjuvant treatment planning. So we can use intraoperative structure updates to estimate how much tumor is left behind and then allow the computer system to do a quick assessment of the quality of a radiosurgical plan to help guide the surgeon intraoperatively as to how much to resect and when enough has been resected when we can stop and avoid the morbidity of taking out that last 5%, 10%, 20% of tumor.
So I wanna give you examples of how we use this to tackle the different areas of inoperative planning and procedure that I mentioned. And I wanna start with the idea of minimally invasive surgery and approach selection. So here's a case of a patient of mine with a small to medium-sized plain sphenoid meningioma, but that was associated with significant frontal lobe edema who presented with seizures. And, you know, you could argue to treat this with radiosurgery, but because of the edema and because of the patient was symptomatic from recurrent seizures, we felt that it was appropriate to resect this.
So this patient, of course, like all patients wants a minimally invasive but excellent outcome surgery. And so how can we perform that? So there is a standard bifrontal craniotomy that we can use to approach this, but there are some better techniques that we have surgically today. One option would be to approach this endoscopically through a transnasal approach, which is actually an excellent way to get at that tumor. But this patient has intact olfaction and this approach almost always sacrifices olfaction, because you have to go transcribiform. And so there are also super territorial or, you know, approaches from above the skull base that can allow us to save the olfactory nerves.
We can approach through a minimally invasive supraorbital craniotomy with a small incision in the eyebrow and do a keyhole craniotomy or we can sort of extend that into a larger cranial orbital keyhole craniotomy that involves removing the orbital bar. But how do we know what the right approach is? Well, we planned a cranial approach because of her intact olfaction and we planned an eyebrow incision for this patient. And here you can see we're looking through the microscope at the patient. The proposed incision site is marked, and we can superimpose the tumor object from the navigation onto the patient to really have a sense of where that is in 3D and what the trajectory is to it.
And you really can't tell what you're gonna need until you get a little bit further into the surgery. But when we actually open that eyebrow and take a look here, we see that, you know, the simplest approach would be a supraorbital craniotomy, but really that would give this approach which doesn't get us down to the base of the tumor, doesn't allow you to get that dural base and achieve a real Simpson grade one or two resection. If we do a larger cranial craniotomy that really gets us down all the way to the base of the tumor.
And so you can see very quickly, it becomes apparent what approach we need, and we do this before we make the craniotomy so that we're not struggling afterwards. So this is, you know, what we achieve. We can remove the orbital rim and the frontal bone in one piece. And then if we look straight down our angle of view and navigate off the microscope, we see that the field of view off the microscope gives us a direct shot right to the base of the tumor.
And so that's gonna be sufficient to be able to take out that tumor. Indeed, we're able to achieve a gross total resection. With these minimally invasive approaches patients heal quickly and are able to return home very fast. This is the patient at a two-week postoperative visit, minimal pain, excellent recovery.
So that is how we use approach selection with intraoperative imaging. Now, what can we do with that added augmented reality that intraoperative visualization to improve postoperative adjuvant planning? And this is where we can combine that augmented reality with adaptive hybrid surgery. So this is an example of a patient that I just recently operated on last week with a petrous apex meningioma.
This is a non-contrasted image because this patient has chronic renal insufficiency and can't receive contrast for a contrasted MR which also tells us that if we leave any residual behind, we're gonna have some difficulty assessing the boundaries of that residual tumor. And if we do wanna treat with adjuvant radiation, we're probably gonna have to over target a little bit to make sure that we cover the area.
And if that's the case, then we need to know that we're very safe in terms of our residual away from at-risk structures, such as the brain stem, optic nerve, because this does go into the posterior cabinet sinus there. And so we can identify the objects, identify our plan ahead of time, and then we can go into the operator room. So this shows you the approach through a retrosigmoid craniotomy, and the line of site here that we get from the microscope right to the edge of the tumor. But what you actually see in reality is this.
So when you perform that retrosigmoid craniotomy, you see a small amount of tumor here sticking out between your trigeminal nerve and your vestibulocochlear nerve. And this is your working window. It seems like you have a good view of the tumor here, but if you actually use augmented reality to superimpose the entire tumor boundary over what you're seeing, you realize that there is a much more tumor than the window that you have to work through.
And this is very helpful. Even though we know this as surgeons, you know, there's no surprise here if you're experienced with this. It still helps to have a visual image of the actual boundaries, especially the upper and lower boundary that are sitting directly underneath those nerves. And so this helps us actually perform a more extensive resection to work piecemeal through this area until we've removed nearly all of that tumor. We're not gonna chase it forward into that posterior cave sinus. But once we've achieved what we think is safe and resectable, we can then use the intraoperative structure update that you saw to get an assessment of how much tumor has been removed. And we actually do this sequentially throughout the tumor resection. But I'm just gonna show you sort of the final resection here.
And ultimately, what we've achieved is a resection of everything down to the base of the, you know, petroclival angle and the extension going into the cavernous sinus. And we can superimpose our residual then back into our microscopic view and see what's left. And if we actually perform an adaptive hybrid surgery run, and we see that we're still too close, we can physically see where that area is that is still a problem for radiosurgery, so that we can target that with the operation.
So here's the adaptive hybrid surgery look. This is the initial tumor. Obviously, too big and too close to the brain stem for single fraction or even really hypofractionated radiosurgery initially. But after we've performed our resection, the residual looks like an excellent target, and we know we can stop here.
And then, you know, it was mentioned that the quality of the radiosurgical or the adaptive hybrid surgery information is dependent on the quality of the inoperative structure update that we get from the system. And so it's always important for us to assess how well we're doing here. This shows you the pre and postoperative images with the initial tumor and the residual tumor marked. And we see that the residual tumor actually matches up very nicely with the residual that's there on the postoperative image. And we look at that with every case to show that we're usually within about 5% volumetrically based on what the system suggests we have as a residual and what we actually get on the postoperative MR.
So how else can we use inoperative imaging and inoperative structure update to help us? So sometimes in the case of adaptive hybrid surgery, we want to intentionally leave some tumor behind and we wanna know can we radiate it. Other times, we wanna be very clear that we're removing all of the tumor. And I think that becomes very important in the special case of hyperostosis meningiomas of the skull base.
So the hyperostosis, this thickened bone can be very hard to differentiate its boundaries intraoperatively, and can also involve areas near cranial nerve foramen and make it very difficult for us to access. But it is important to remove this because often it is the hyperostosis in these cases that's causing cranial nerve compression and the patient's actual mass effect and symptoms. And also we know that this hyperostosis harbors live tumor and needs to be treated either surgically or with adjuvant radiosurgery in order to really get long-term tumor control.
So, you know, this is a case of a patient where you can see the hyperostosis in the sphenoid wing actually leads to orbital proptosis, which is both a cosmetic issue, but increases pressure on the optic nerve through the optic canal and within the orbit and leads the vision loss. And if you don't remove this completely, you just remove the soft tissue component here. This patient's vision is not gonna get better after surgery.
And so we can utilize the combination of preoperative MRI because I don't have intraoperative MRI at my institution, but we do have an intraoperative arrow CT, and we can fuse those images, CT and MR, to get a look at not only the soft tissue component of the tumor but also the bony component of the tumor, which is much better visualized on CT.
And then we can use our image painting tools to identify as separate objects, not only the soft tissue component of the tumor, but also the bony component of the tumor. And then combine those two to actually look at the relationships between the hypersonic bone and the soft tissue component. And we can use the intraoperative structure update then as we go through to look at how we're doing with our bony removal, completely separate from the soft tissue component of the tumor. So that see, we started with this large area of hyperostosis, and then we can use the inoperative structure update to go through the bone as we are resecting it, and estimate how much of that hyperostotic bone is left.
And again, when we look at the final outcome, we see that what is shown is the original tumor and the residual tumor matches up very nicely between the inoperative MRI based on the preoperative scan and the postoperative scan, which shows we've actually achieved the resection we wanted to. And so these are ways in which, again, inoperative imaging and these tools can help us maximize the goal of our surgery.
So in conclusion, you know, I think it's important that we utilize these intraoperative image guided navigation tools which can help us plan more minimally invasive procedures when appropriate, it can help us limit the extent of our resection so that we're being maximally safe but also maximally effective in terms of long term tumor control, and occasionally to maximize the extent of surgical resection when that is the goal of our surgery. And the integration of this information with the knowledge that we're gonna treat with adjuvant radiosurgery really helps us to perform optimal surgeries for this combined multimodal therapy. So I thank you very much for your time.