Well, we're going to go from the most common tumor to the least common tumors, I think. Thank you, Mike and Manish, for inviting me. And we're going to talk about if I can get my little mouse to work here. Thanks. Rare tumors, the malignancies of the skull base, but important to recognize and at least set the patient on the correct course of management. This is about a year old, and the numbers aren't important, but just the relative percentages are. There's a real difference in the nature of tumors between what you see in children and in adults. The adults are predominantly epithelial malignancies, predominantly from the sinonasal cavities, and sarcoma is a less common variant, although in our institution, which has an emphasis on sarcoma, we do see about 40% of those in adults. Children is primarily sarcomatous tumors with epithelially originating tumors distinctly rare in that group of patients. And the distribution also is different, but in adults and in children, in the adults, you can imagine in the anterior skull base, the epithelial or, you know, the sinonasal tumors are the most common. They decrease as you go posteriorly, as there's less epithelium posteriorly. And as you see, both in the middle cranial fossa and the posterior cranial fossa, the soft tissue and bone tumors are more common. This is distinct from children, where sarcomas are more common at all three skull base sites, and especially in the middle cranial fossa. This is the nature of tumors and malignant tumors in the anterior skull base in adults. You see sarcomas make up almost a quarter, but then squamous cell carcinoma, olfactory neuroblastoma, also called asthesial neuroblastoma, adenoid cystic carcinoma, adenocarcinoma, and sinonasal undifferentiated carcinoma make up the majority. These are the various sarcomas in the anterior skull base. This is the middle skull base. Sarcomas over 50%, squamous cell, and adenoid cystic get there, usually through the trigeminal nervous system or direct extension out the back of the maxillary sinus. Sarcomas in the middle cranial fossa are a little bit more diverse, including hemangiopericytomas and rhabdomyo-sarcomas. And in the posterior skull base, the majority are sarcomatous tumors, chondrosarcoma and chordoma being predominant. The reason I showed all those is that was 35 different tumor pathologies, over 20 different distinct varieties of sarcoma, and over 15 non-sarcoma pathologies. So it's clear that the management of these tumors has got to begin with a multidisciplinary team of pathologists, diagnostic imagers, surgeons, medical oncologists, radiation oncologists, and all the support people that are necessary in order to optimize patient outcomes. There are things that I think are important when dealing with someone with a malignancy of the skull base. I think imaging is critical. Biopsy is the only thing that allows you to do the right thing intellectually with respect to management pathways. Expert pathology review, the discussion in a multidisciplinary team setting, and finally surgery and its role within the multidisciplinary plan. So the CAT scan and the MRI are complementary studies. They both give you distinct, important information about the tumor. Sometimes they guide you to the diagnosis. Other times they help for surgical planning. PET scan is very useful, both in identifying distal metastasis and for also grading response of the tumor to your adjuvant therapies. Imaging, you want to do post-contrast, fast suppressed imaging. That allows you to see the differentiation between a contrast enhancing tumor and the surrounding fat of the marrow and of the skull base, the orbit. T2-weighted studies are particularly important for chordomas and chondrosarcomas. MRV is important for tumors involving or abutting the major venous sinuses. And of course, CT and CTA are important for visualization of bone in the skull base and for structures, vascular structures passing through the bone. Here's an example. This is a sphenoid wing meningioma, not a malignancy, but the CAT scan clearly identifies the hyperextotic bone that requires resection. This is an osteosarcoma showing ossification and osteoid deposition in the sphenoid sinus and clivus. The olfactory nerve blastoma, this is a nonspecific imaging. It just shows a destruction of the ethmoid air cells and of the medial maxillary wall. And here's a post-contrast, fast suppressed imaging, showing the tumor in the orbit there on the left. If you were not fat suppressed, you wouldn't see that tumor on the right because the fat would be bright. So I think fat suppression is important. This is a chordoma on the post-contrast T1 axial on the left. You can barely see it. It's that little stipple enhancement there. Clearly much better identified on the T2-weighted imaging. And here's just a meningioma with absent sinus. Olfactory nerve blastoma, you can see how CT and MRI are complementary. Here's a plasma cytoma. CT shows loss in the petrous apex and the clivus on the left side. It is contrast enhancing homogeneously. I don't have the T2-weighted images, but the differentiation here is this is a non-T2, hyper intense tumor, unlike chordoma or chondrosarcoma, which would be T2 hyper intense. And here is a PET scan in a patient with a giant cell tumor identifying response to a denosynab. Biopsy is really important. Many times it can be endoscopic if the tumor presents in the side nasal cavity, but sometimes for the soft tissue tumors in the middle fossa or the posterior fossa, CT or ultrasound guided biopsy is critical. Rarely you have to do it open. It's only with pathology in hand that you can slot patients in these kind of different management paradigms. And what's important is not only to do the biopsy, but to get the right people looking at the biopsy. You really need an expert review. You need expertise in neuropathology, head and neck pathology, bone and soft tissue, sarcoma and hemopathology. All these tumors can occur around the base of the skull and they have distinctly different management pathways. In a study we did, Dr. Jan Bruner looked at 500 specimens and changed diagnosis in about 42% of them. The change in diagnosis was serious in 8 or 9%, substantial in 20% and minor in 10%. We also looked at this in the sinonasal cavity. These are three different tumors, three distinct tumors, three tumors with very different management pathways, olfactory neuroblastoma, neuroendocrine carcinoma and sinonasal endocrine carcinoma. We found on review of 12 consecutive tumors all called, I think 12 or 10, I can't remember. Those numbers, all called olfactory neuroblastoma. When we re-reviewed them and included immunohistochemical studies, only two were actually asthesia neuroblastomas. Two were actually melanoma and a whole bunch were benign. So some of these benign tumors were treated so high that they had bilateral optic nerve radionecrosis. So I think both overdiagnosis and underdiagnosis are common and need to be guarded against. And then once you have the pathology in hand, you get back into your room there with your colleagues and all these disciplines and you look at these different paradigms. You know, the simplest paradigm is surgical resection. There are definitely tumors that this is probably all that's necessary. Some low-grade chondrosarcomas, grade 1 chondrosarcoma or a totally excised grade 2 chondrosarcoma may not require any further therapy. Similar basal cell carcinoma can be cured with an aggressive resection. Most patients still are managed by surgical resection and postoperative radiation therapy at our institution. But an increasing number are entering now multimodal plans of neoadjuvant chemotherapy followed by either surgery and chemoradiation or different combinations thereof. And I'll get to that in just a second because that's how things are changing nowadays. So this is the importance of a multidisciplinary team. This guy came in to see us with this tumor and an acute neurologic deterioration. You can see he's hemorrhaged around this tumor in his brain. And this guy was kind of, he was going south pretty quickly. And I, you know, we really didn't have a whole lot of time to create a complex multidisciplinary plan with multiple surgeons. So he went to the operating room and I did a craniotomy and took out his intracranial portion of that tumor. I left the sinonasal portion of that tumor alone. There's his post-op scan in the top panel. He then was diagnosed with a sinonasal undifferentiated carcinoma. He got four courses of toposide and cisplatinum. You see he's had a partial response. Interestingly, he had reossification of his ethmoid bone. But he still had visible residual disease, which we went ahead and did a formal craniofacial resection for and then treated him with post-operative radiation. He is alive now actually 10 years later. And his pathology identified not a single viable tumor cell. So I think this is why the guy's alive, is the chemotherapy essentially sterilized his tumor. And here's another great example of a squamous cell carcinoma in the cavernous sinus along the trigeminal nerve, treated with carboplatin, paclitaxel, and gemcitabine, post-op radiation therapy, alive three years later. So I think that this is not a surgical disease that was cured without it. This is just the point of this slide is that there's over 3,500 patients at MD Anderson from the 1940s to present. There's been an increase in the multimodal nature of their plans, not simply radiation or surgery. It's becoming increasingly complex. And I want to use this example, the sinonasal endopharyngeal carcinoma, as a paradigm for this. So we have changed how we manage this disease. If you don't know about sinonasal endopharyngeal carcinoma, it's viciously aggressive. Its survival profile is worse than glioblastoma almost. So we've gone from surgery to starting almost all our patients now with induction chemotherapy. And then following the patient's response to induction chemotherapy comes the decision of whether to go on directly to concurrent chemoradiation or to surgery. And what we did, you know, this is the numbers. And you can see they all got induction. One group either had complete response or partial response. That was 64 patients. The other group had less than a partial response. That was 31. And this is how they were subsequently treated. These are our definitions of complete, partial, stable disease, and progressive disease. So first of all, right off the bat, we see that with this management paradigm, we're looking at five-year disease specific survivals of 60% for sinonasal endopharyngeal carcinoma. Massive increase over historical survival data. But this is what's really interesting. When we look at the overall survivor, those that respond to the chemotherapy have a highly significant improval in their overall survival and in their disease specific survival. Then when we looked at the responders, this is non-responders. If the patient responded and we then treated them with chemoradiation, and that's the top left panel showing a significant improvement in those patients that got chemoradiation rather than surgery. And that's the same in disease specific survival. In fact, the people who got surgery after their induction chemotherapy did very poorly. And if you look at the non-responders, it is in this group, the patients who did not respond to induction chemotherapy that surgery did have a role in improving their survival, although it was generally poor to start with. The other thing we think is happening is if here's a disease free survival with chemoradiation versus surgery and chemoradiation, but then local regional recurrence and distant metastasis. What our hypothesis was is that in those patients who got surgery first, it was a systemic disease and not a local disease. And by using the chemoradiation, we were preventing or decreasing the incidences of both local regional recurrence and distant metastasis. So about two-thirds of our patients responded to induction chemotherapy. Those that responded had a better role of prognosis. Those who responded and treated with chemoradiation had good disease control, good organ preservation, and improved survival. And non-responders to induction chemotherapy were best treated with surgery if resectable, then followed by chemoradiation. So that's a different paradigm than what we're used to. And that's going to be more and more important as we move forward as neurosurgeons to be part of this. What we want to do is we want to keep on getting better. So any of these new approaches, you know, have to hold the gains that we've made over the past 40 years. And when we look at our current standards, this is what I define current standard. It has to be MR era, and it's about 1990 on. This is our 3500 patients looked at and stratified by decade. And you can see we have been making incremental improvements right from the 40s to present. And you can see what our survival of three years and five years now is from the 1940s into the 2000s. And what our histological outcomes are over that same period of time, over 10 years now. So they've been improving. This progress is attributed to, you know, first of all, most effective, more effective surgical treatment, which was made possible by the advent and subsequent refinements of craniofacial and skull base resections. More effective reconstruction using vascularized flaps. These patients aren't falling apart and dying of infection. More effective adjuvant therapy, particularly conformal radiation therapy was another big plus. And now we're in the era of chemotherapy adding to our survival outcomes. And we are on the verge of entering into the targeted therapy era to help us with some of these very difficult diseases. So again, this grouping of people is going to become increasingly important in what we do. So one of the big controversies in surgery of sinonasal malignancy especially has been whether or not endoscopic techniques are a violation of the concept of unblock resection and whether or not that's, you know, we're doing bad things. And we looked at this over the years. Now, this is our follow-up data, which we have not published yet, looking at now that number. And what we found in well-selected patients that there is no difference in outcome whether you do this operation purely endoscopically or endoscopically with cranial assist. So this is just where all those tumors were. But this is a kind of proof of principle that in the well-selected patient, it's not so much the surgical approach or surgical technique. It's the achievement of margin of negative resection, which is the most important thing. And now it's more, which is just the better approach. This is clearly, both these tumors are better done endoscopically. There's no reason for considering cranial approaches in this kind of a patient. These patients, however, are not well done endoscopically. You know, there's frank orbital invasion where the orbit probably has to go oncologically. Extensive, both midline and lateral brain invasion. Or very, very anterior disease, which is a blind spot endoscopically. These are all cases where a combined cranial endoscopic approach is probably the best thing. And sometimes, just for palliative, this is just reestablishing an airway in this patient with a very extensive mucosal melanoma. What has happened over the last three decades, and I kind of feel a little sad that I'm talking about three decades of practice. That we have moved from the classic bifrontal craniotomy, Weber-Ferguson cranial facial resection, to an era where we've been trying to, we're trying to avoid any kind of facial operations. Did a lot of purely transcranial operations. And that was the 2000s with the birth of the endoscopic approaches. And now you can see what we're doing now in the 2010s where there's a significant amount of our patients are managed purely endoscopically or cranial endoscopically. This is a perfect paradigm for this. This is a olfactory neuroblastoma. There is a little bit of extension over the orbit there on the dura on the left side. So this is a tumor that's best managed by cranial endoscopic approach. We would start endoscopically, you know, in order to leverage the improved visualization of the sinonasal cavity. And then resect the intracranial component. This is your standard bifrontal craniotomy with the pericranial flap reconstruction. And what you see here is the look from above where we're directly suturing the defect in the dura. We have the robust pericranial graft about to be laid in. And there we are looking from below. We can see our reconstruction from below. So this is really we found to be the best of both approaches. Of course, it avoids facial incisions. It leverages the superior optical navigation of the sinonasal cavity afforded by the endoscope. Gives you a panoramic, unobstructed view of the entire floor of the anterior skull base. Gives you the ability to manage dural defects by direct watertight enclosure through a transcranial approach. Gives you robust, reliable reconstruction with vascularized pericranial tissue. And then gives you improved options for orbital, cerebral, and frontal sinus extensions of tumor. And at the end, gives you direct visualization of the adequacy of your section, of your reconstruction. When we, over the years, have looked at the traditional surgical contraindications to surgery of malignant skull base disease, these were what have been traditionally listed as relative contraindications. We tried to, you know, look at each one of these individually and systematically. First of all, we looked at the transdural spread of malignancy in this group of patients. All of whom had spread into the skull base through the dura. Our survival in the modern era was about 67%, which is about the same as those patients without intracranial extension. And the most important finding was what I mentioned earlier, the ability to achieve a margins negative resection. And it did not matter if you did that open or endoscopic, or piecemeal, or en bloc. So I think that this paper, I think, cemented the concept for us that it doesn't really matter about the technique as long as you, at the end of it all, get margins negative resection. Perineural extension is a disease specifically linked to adenoid cystic carcinoma. Also can occur in desmoplastic melanoma, and squamous cell carcinoma, and several other malignancies. Here's a maxillary sinus adenoid cystic with extension into the maxillary division of the trigeminal nerve. We, you know, have extended our extent of resection by following the trigeminal nervous system intracranially, following a standard maxillectomy for the resection of the primary disease. And through that approach, you can access the intralateral cavernous sinus. You can follow the maxillary nerve in. You can actually reach the caesarean ganglion, like you see in the bottom right, in order to resect that part of the tumor that's heading backwards. We found that important because in our series of patients with adenoid cystic, a microscopically margins negative resection imparted a significant survival advantage compared to leaving positive margins behind. So we have been fairly aggressive in pursuing that type of disease. The infratemporal fossil was considered sort of a predictor of a bad outcome, and given its location and its proximity to the para, the parapharyngeal and carotid space and middle cranial fossa, anterior cranial fossa, posterior cranial fossa. This is a chondrosarcoma. These can be resected too. And when we looked at our patients with intralateral malignancy, and 39 of those 52 patients had infratemporal fossa extension. All these had multimodal management. Our median overall survival in these patients was 4.7 years. And our five-year overall survival was 53%, which is in keeping with other areas. And so I don't think it is specifically a negative prognostic indicator. I think we have the abilities surgically to manage these patients. So I don't think that's a problem, but this is a bit of a problem. Patients with carotid artery involvement, either in the temporal bone or the cavernous sinus, in this little paper we dissected the tumor off the carotid in four patients, and we cut the carotid out in four patients. And, you know, great. Local recurrence was more common where we dissected the tumor off the artery than it was if we just cut the artery out. But the overall median survival of patients not requiring cavernous sinus dissection or carotid resection was twice that of patients undergoing these measures. Even though we cut the tumor out, cut the artery out, there was less local recurrence, they all died of regional and distal metastatic disease. So we're not dealing so much with an anatomic problem, we're dealing with a biologic problem. And I don't think we currently have the ability yet to manage this. And we don't advocate surgery for high-grade malignancy when in cases the cavernous carotid or the petrous carotid. We've had the opportunity to operate on patients with metastatic cancer to the skull base. It is a rare event that we do this. Majority of these patients are managed effectively with radiation therapy, specifically highly conformal radiation therapy techniques. But you get stuff like this when a patient is losing vision in both eyes and had a metastatic glioma and sarcoma. You know, you've got to do something. So we went ahead and resected that transbasal. Nowadays we would have done this endoscopically. This was quite a while ago. There's the two optic nerves post-op. So she died 29 months later of disseminated disease but had no local recurrence and no visual difficulties for the rest of her life. So when we looked at this group of patients, it was only 27 when we published this paper, there was a distinct reason for operating these patients. There was a distinct problem that we were trying to palliate. And what we did here, we put down, this is where they're from. What we did here, this is the indication for surgery. Progressive optic neuropathy, progressive perptosis, diplopia, increasing mass or solitary metastasis, pain. And then at the end, the outcome is whether or not we did achieve palliation. And, you know, we did pretty well most of the time. Achieved that. But we didn't have any surgical mortality. We gave one patient a new sixth and a new third transiently. Their survival is 11 months, which is about the same as when you get multiple brain metastases. And all patients, like those patients with multiple brain metastases, die of systemic progressive disease. But to look at these last little things, there are patients that you should select out who do very well. A patient was over, survival was over two years in five of the seven patients with sarcoma. And survival over nine years in patients with metastatic follicular thyroid cancer. So I don't think that these patients should be dismissed. They should be, again, discussed in a multidisciplinary fashion. And they are, should be candidates for surgery if that's deemed important and necessary for palliation. The elderly are getting, there's more of them around. It's growing rapidly. And advanced age may be considered a relative contraindication to aggressive resection. But the elderly population is healthier than the past and can expect a mean of 17 further years of life once the age of 65 is reached. So that's also an important group of patients. The punchline of this paper was that there was no oncologic difference in outcome either between the young age group, which is a median age of 56, and the elderly age group, which is a median age of 70. And this is, one thing you did note here is there are, there is a threefold increase in the incidence of systemic complications in the elderly group. They come to you with a bunch of pre-mortem stuff and hypertension and smoking. So I think our, what we decided from the outcome of that paper is that yes, those patients are important patients to be considered for surgery. Yes, they have more complications. Yes, you have to be particularly careful in their perioperative care in order to achieve the benefit of an improved oncologic outcome. So here's those lists again that we talked about. We've been left with these two with poor answers for. Cavernous sinus and internal carotid artery involvement and melanoma histology. We have looked at melanoma a little bit more carefully because this is an interesting, very interesting tumor in the era of targeted therapy and immunotherapy. And overall, 41% of our patients with sinonasal mucosal melanoma harbored genetic mutations. The BRAF and KIT mutations were identified in unfortunately only 8-5%. And patients with, let me just see, so margins are important in this group of patients. Unfortunately, mutation status didn't show any associated survival outcomes. We did find NRAS mutations in 30% and they were, you know, they're interesting because they can be targeted by MEK inhibitors. So that's the next phase. What we're doing now is trying to tease out these, the NRAS mutations and put them on trial with MEK inhibitors and obviously our BRAF patients on BRAF inhibitors. But, and we haven't talked so much about immunotherapy, but that's also being looked at more carefully. We just haven't, the data yet. So to summarize things that I think are important, both CTMR and PET are important for staging and response. Biopsy, use the most accessible route. And if you do endoscopy, endoscopic biopsy, do it in the operating room. Don't, please tell your head and neck colleagues not to biopsy patients in the clinic. Pathology review is critical. It is critical to the formation of a treatment plan. It is critical that it be reviewed by experts. Then take that information to your multidisciplinary team and talk it out, figure out what trials are available, what order we're going to do those different things in. And this is important in order to form and deliver the treatment plan. And surgery is going to have a significant continued role, but the role is evolving. And there's going to need to be a reevaluation of our timing, our extent and goals of surgery. All that will be necessary to reevaluate. And the neurosurgeon needs to be actively part of this in creating the multidisciplinary and multimodal plan. Because, you know, if you're going to get three courses of chemotherapy, you've got to know when that patient's going to nadir so you know when to put them on the operating room in order to resect them. And patient selection remains critical in determining the optimal surgical approach for these very challenging neoplasms in these patients. Again, the best patient outcomes are achieved when the best quality surgery occurs as part of a carefully constructed multidisciplinary management plan. Thanks very much. Thank you. Thank you. Right. So if there is no effective adjuvant therapy and you can't get the tumor out completely in a high-grade malignancy, there's probably no reason to operate on that patient. You know, you send them down to palliative radiation therapy or maybe some combined palliative chemo radiation. Because their problem is beyond what a surgeon can actually help them with. Those patients we did the palliative surgery in the metastases, you know, we were still achieving, those patients were selected for a total tumor removal. So we weren't doing, kind of, palliative, little sub-tumor resections on those people. There's, like, you know, the surgical oncology or the surgery we often do, aren't as part of the surgery or the surgery in the brain. Yeah, this is completely different. You know, this is, I think, the whole endoscopic thing is different, too, between operating endoscopically in a meningeal marpitutary enoma and operating endoscopically in a malignancy. You just, those are two different things and have to be philosophically managed in completely different ways. I would say that we very much jumped on the bandwagon also for tumor resendings and differentiated carcinoma. We do up-front chemo radiation and then often resect and then follow-up by, sort of, consolidation treatment. And there's, sort of, two scenarios that I wanted to ask you about. So I've run into where the patient gets 53 up-front with chemotherapy as a partial response and trial for a resection, thinking we can likely get it in a large vein. Recession, we don't, we can't. And then what do you do? Because in my way of thinking of it, we burned a bridge. Because if you had done that as a first step, you could come back and treat them with, let's say, 60 gray for microscopic disease, which I think has a much higher chance of being effective. Considering when you do 10 more gray, that's not going to be the answer. Have you run into that in, sort of, what's the philosophy? Right. So, yes, we have run into that. In those patients where we've, usually those patients have responded to induction, have an income, a good response, but incomplete response. They go on to chemo radiation. They, and they get, again, an excellent response, but have visual residual disease. You operate on them and you leave some microscopic disease. Those patients we re-irradiate, ideally with focused radiation techniques. Because usually you can sneak in a little extra 10 to 12 gray, maybe 14 with a gamma knife in those patients. So, we try to use those focused techniques. And then it becomes a risk benefit, sort of, scenario where you have to decide whether or not re-irradiation to higher doses is worth the risk of radiation injury to the brain.

rare tumors

skull base

multidisciplinary team

diagnostic imaging

surgical resection

multimodal treatment

research