Hello everyone. My task today is to speak about interoperative ultrasound in high-grade glioma surgery. The mainstays of image-guided surgery are based mainly on conventional neural navigation, interoperative MRI, fluorescent guided surgery, and interoperative ultrasound. Let's shortly review the pros and cons of each of those systems. For instance, conventional neural navigation is certainly based on standard imaging. Its use is now routine in almost every operating room. It can be also integrated with more advanced image systems, for instance, with more advanced segmentation of the images, which can be coupled to the interoperative navigation as well. They do not serve only to the purpose of pre-planning of surgery. However, there are some drawbacks. Those are based on the fact that this imaging is based on preoperative imaging, therefore cannot be considered a true real-time navigation, but rather a virtual navigation, and it does not take into account the brain shift phenomenon. Interoperative MRI has been regarded as the gold standard of interoperative imaging because it's based on standard imaging. It's extremely accurate. Here you can see some images taken from our operating rooms related to some cases. The quality of the image is outstanding. However, there are drawbacks. Those are related to the fact that it requires a dedicated area and tools. It is extremely time-consuming. It is extremely expensive, and on top of everything, cannot be considered a true dynamic online imaging system since you have to interrupt your procedure, transfer the patient into the scan, get the exam done, and then take the patient back to the operating table. Fluorescent eyes, for instance, fluorescein, on the other hand, are real-time interoperative image system. It is a fluorescein, a particularly inexpensive, however, it marks the interstitial space, therefore lacks of specificity. 5-ALA is real-time, marks tumor cells, and in that it's true, specific. However, it works only on high-grade gliomas, and also even a thin layer of any kind of fluid, can be blood, can be saline, may hamper the visualization of the fluorescence. Interoperative ultrasounds, on the other hand, is real-time, is true real-time imaging. It is dynamic, you can do that, you can use it at any time during the surgical procedure, and it is relatively inexpensive. And the interest among surgeons is raising through years, as you can see from a growing body of literature, especially during the last few years. This is due to the recent dramatic improvement of the definition and resolution of the technique. As you can see, you can even distinguish between the gray matter and the white matter in this scan. However, there are also downsides. Based on the type of imaging, which is unusual to us as neurosurgeons, and also there are difficulties due to the orientation, for instance, of the probe. We are accustomed to look at the MR images or CAT scan images on the three orthogonal plants, and here you have an infinite way to slice and scan the brain. Therefore, this makes of this technique a strictly operator-dependent technique. And as a matter of fact, even for an experienced neurosurgeon, it can be very difficult to identify all the anatomical landmarks, which you can see depicted in this scan, unless you don't have a coplanar MR, which indicates clearly that this is the tumor, this is the dentorial edge, this is the cerebellum, and this is the choroid plexus, which is always hyperchoic, this is the brainstem, which is always hypoechoic, and this is the pineal gland, which is always hyperechoic. Those are basics of the semiotic of ultrasound. Here is another example where you can clearly see that this is the tumor, this is the dentorial edge, this is the farx, and this is the pineal gland. Here, again, another example where you can, if you have the coplanar MRI, clearly identify the tumor over here, you can see the farx, you can see the dentorium, you can see the cerebellum, here is the fourth ventricle, again, this is the straight sinus, and you can even see the hyperchoic choroid plexus of the contralateral ventricle. Here, how we use it, usually we use it, we do the first scan once the skull has been opened, the dura is still intact, in this case, you can see very well the tumor, the cystic component, you can see the choroid plexus hyperchoic over here. Here is another example, you can see the tumor, the cystic component, and you can see very well the farx, the straight sinus, and the dentorium, and the cerebellum. Then, one thing that you can do, you can adjust also for the brain shift, here you can see the two images are not overlapping, so you simply freeze one of those, in this case the MRI, and then we drag it over the ultrasound image, and you do, and you repeat this operation several times in order to adjust for the brain shift. Here, again, another example, here you can see that the farx, the two images, the virtual image, and the real image of the ultrasound, they do not overlap, in this case we freeze the ultrasound, we drag it over the virtual image, and in this way we adjust for the brain shift. Then, you can use ultrasound at any time during surgery, for instance, in this case, just to assess how much of the tumor you have removed, you see that only the tumor has been removed partially, this gives you the exact perception of how much of the tumor is left over, and how much work you must keep doing. Here is the end of the resection, where we check the cavity here, those are all air bubbles due to the irrigation, and we check on the borders of the cavity for tumor remnants. Here is another example, you can see very well the two ventricles, deceptive perlucidum, here we're using a convex probe, which has been put in contact with the bottom of the resection cavity, in order to avoid all the artifacts that usually are at the interface between the fluid and the resection cavity as well. You can also use more advanced MR imaging, such as DTI or functional, here you can see in bluish, depicted in bluish, the reconstruction of the motor pathway, here you can see the arcuate fascicle, and for those of you who are interested, this is an interesting article to read. Then, there are additional features we can exploit with the ultrasound, for instance the use of Doppler, elastosonography, and contrasting, and use of contrast, for instance. Here you can see the use of Doppler, it's quite straightforward, I would say, those are the pericallosal arteries, and here you can see the callosomarginal artery, which was running inside this tumor, the tumor has been removed, this is the resection cavity, and we check on the patency, for instance, of the callosomarginal artery. Elastosonography, it's useful in order to understand which is the consistency, the stiffness of the of the tissue you are scanning. You see that is a referring scale over here, which is a color scale, and in this case, red is hard and green is a soft, those have two different techniques, but I won't spend time on this because it's too complex, and here you can see how we visualize the tumor, which is, in this case, is quite hard as compared to the rest of the of the brain parenchyma. Here's a video clip where you can see that the tumor is quite hard, is reddish, and the rest of the brain parenchyma is softer, and here is another example, the opposite, this tumor, which was a metastasis, is quite soft as compared to the rest of the of the brain parenchyma, and of course you can use also DTI coupled to this kind of images, and here you can see some example of elastosonography done with different kind of tumors, but more interestingly, the elastography can be used as a useful tool in detecting the remnants along the borders of the dissection cavity, and those are other examples which have been published in this paper quite recently in Operative Neurosurgery, which basically confirmed the data published previously by Selbeck and a group of Norway led by Uzgari in 2005, 10, and 2012. The use of contrast. Contrast is useful in order to enhance the anatomical landmarks, give us some hints about the histology of the tumor, we can perform angiosonography and identify tumor remnants. So let's see how. In this case, you see this glioblasoma with cystic components, parenchyma components, you can see the feeding arteries over here, the uptake of the contrast, of course the necrotic part, the cystic necrotic part is not showing any uptake, and as you can see here, draining vessels will appear in this phase toward the end of the contrast administration. Here again, you can see a tumor with a huge cystic component, this is the feeding artery, and this is the uptake of the contrast through time. Here you can see a frontal tumor, those are the pericallosal arteries, you can see over here, and then you will be showing at some point the appearance of the callosomarginal artery, which is this one, let me show you, this one over here, and this information was important in order to understand where the artery was located in order not to endanger this, and as a matter of fact, we checked it towards the end of the operation if the artery was padded. This is the same case of the Doppler that I showed you earlier. Histology, we know that according to the grade of gliomas, there are different degrees of uptake of the contrast, with glioblastomas showing the highest level of uptake. Interestingly, in radiation necrosis, we never found any uptake of contrast in the borders of the cystic lesion, as opposite as cerebral abscesses and colon metastasis, and here are other examples. Androsonography, you can even perform an androsonography, this is a clinoid meningioma, but I showed you this because it's very interesting to how we can identify the polygynous willis, as you have seen over here, and the polygynous is clearly depicted, and this information is extremely important during surgery in order to understand the anatomy, and this also gained the cover of Acta Neurochirurgica a few years ago. It is extremely useful also, the use of contrast, in order to confirm the hypothesis of some fragments of tumor left over, some residuals, as you can see on those images. Here is an example, this is the tumor before the section, and here you can see the the adjustment for the brain shift. I go quickly over there, then we perform the use of contrast in order to understand better the vascularization of the tumor. We do elastosonography to assess the stiffness of the tissue, and finally we check for tumor remnants, and this is how we check and we recheck again after removing some parts which were looking like possible residual tumors. And here are some static images, you can see the collapsed cavity over here, and then you can see some hyperechoic signal here and here, in the posterior aspect of the cavity, suggesting the possibility of a residual tumor. So we administer the contrast, we see that there is some uptake of the contrast in the posterior wall of the resection cavity, and at the bottom of the resection cavity. So we went back, we removed those parts, and in this way we achieved the complete removal of the tumor. And for those of you who are interested, those are some interesting articles, but more importantly, based on those observations, the European Federation of Ultrasound Society has introduced as a specific recommendation the use of the contrast in neuro-oncological procedure for tumor identification, for the assessment of the boundaries of the tumor, and for the evaluation of residual tumor, as well as also the recommendation to use contrast to do angiosonography. And for a more extensive review of what I just said, you may want to check out this book. Now, those are the mainstays of intraoperative imaging guided surgery, and the question is, can these techniques act synergistically? Now, there are no evidence-based demonstrations. It is a very interesting paper showing the synergy deriving from the use of 5LA and intraoperative MRI in achieving a complete resection of tumor, but nothing has been done on ultrasound. Here is a case. This is a right temporal cipital and a GBM. Here you can see the tumor, FACS, stent notatorium. Here is the cerebellum. We go ahead with the resection. We always use 5LA. As you can see, during the resection, we stay at the boundaries of the tumor, at the very interface of the limits of the tumor with the healthy tissue, which is showing no enhancement. We automated the use of the ultrasonic aspirator with the bipolar, as you can see here. And at the end of this removal, the impression that I got was that I had removed completely the tumor, as you can see here, okay? But then we went back with the ultrasound, and with the ultrasound, we were looking at some residual hypergogenicity over here. So we did the contrast, and as you can see, there is some contrast uptake. Now, notice that this is a dynamic exam, the ultrasonic. So you have to do that moving the probe continuously in order to understand if there is something that consistently is showing uptake, as you can see here. So there is the doubt that there is some remnants, and as a matter of fact, after scratching the tissue, we observed some residual fluorescence underlying maybe some healthy tissue or maybe some fluid which was hampering the visualization of the 5LA. Of course, this is very anecdotical, and there is a strong need for additional trials to assess the impact of combining different intraoperative imaging modalities in achieving better results in removing the tumor. As I said earlier, this is a strictly operator-dependent procedure, and this requires a specific training. Therefore, we have developed a new device and app for the rehearsal of intraoperative ultrasound. So this is a true simulator. I'll show you how it works, as you can see here. Basically, this is based on an app which is downloaded into your smartphone, then the smartphone is paired with the computer, and we have a number of cases you can browse through. For instance, in this case, frontal tumor, we are rehearsing, we are simulating the intraoperative scan, having the two images, the ultrasound scan and the coplanar MRI. As you can see, you can use all the features of the system that we use. You can fuse the images, you can fuse the MR images, and the ultrasound scan in order to better understand the semiotic of ultrasound. And we've done studies with residents, and we found that, for instance, at the beginning of the training, they are able to identify, using the ultrasound, approximately 60% of the anatomical landmarks that they were requested to identify. And at the end of the training, they were able to identify 100% of those anatomical landmarks. And here we come to the conclusions of the presentation. I hope that I gave you clear evidences of the fact that intraoperative ultrasound, enhancing the conventional neuroimplication, is reliable, is accurate, is safe, it's easy to use, can be repeated as many times as you want during the surgical procedure, giving you a continuous real-time feedback, allows also for the compensation of brain shift and may increase the extent of resection. It can be also implemented with more advanced MR imaging, such as functional MRI or deficient sensory imaging, DTI. On the other hand, intraoperative ultrasound is operator-dependent, requires training, and the learning curve is quite steep. And to this end, the use of simulation of a simulator, such as the sim that I showed you, may represent a formidable tool to understand and recognize the semiotic of intraoperative ultrasound. And now I would like to acknowledge all my collaborators in my department at the Neurological Institute Carlo Pesta of Milan, and I thank you for your attention. Thank you very much.

intraoperative ultrasound

high-grade glioma surgery

image-guided surgery techniques

real-time imaging

operator-dependent nature

tumor resection