Method One: Avoid Injuring Them

This is an obvious piece of advice, which is harder to execute in practice. The easiest arterial hole to repair is the one you don’t create. Bleeding is inevitable in glioma cases, as is the need to stop bleeding numerous times during the case. I suspect that many surgeons probably aren’t aware at the time that the bleeding they are stopping is from a critical artery as they are doing it. Thus it may not dawn on many people why they got a bad outcome from a particular case, especially if the infarct is relatively small (but critically located).

While it is possible to perform glioma surgery without a microscope, it is also possible to perform it with the naked eye, and there is essentially no way in which it is better to have less light and magnification. Many neurosurgeons believed in the past that loupes and a headlight were adequate visualization for pineal and aneurysm surgery, an idea which has since generally been dismissed as substandard practice. Today, the same arguments are still put forward for glioma surgery as if they are valid, namely that there is no anatomy to see in these cases. Yet when Drake brought the microscope into his basilar apex aneurysm surgeries, he suddenly saw the perforators and preserved them, making an inoperable problem treatable. It is amazing to me that people maintain the cognitively dissonant argument that certain gliomas are “inoperable” yet you don’t need to see any specific details like small vessels and subtle sulcal borders, which allow many of these tumors to be removed. It is shocking that we cling to 19^th century technology so tightly.

I do not allow my residents to use loupes in these cases, as I believe that the use of loupes to remove a glioma encourages surgeons to attempt things that they would do better with better visualization and light provided by a microscope (or other similar visualization tools), and provides this confidence in the absence of being able to see the critical arterial anatomy (or basal ganglia anatomy, or sulcal anatomy, etc.) that could doom the case if you are not aware of it. Thus, while people can do glioma surgery without the microscope, I have yet to hear a coherent reason how it is better done with loupes, or how a microscope make the surgery worse. If it were your brain, you would probably want your surgeon to have the best available visualization, and that most certainly is not loupe based technology.

For similar reasons, I do not use ultrasonic aspirators in these cases. Beyond just cost considerations, these devices were not developed for intra-axial brain tumor surgery, and it is exceedingly rare that these tumors have a fibrous character where a powered aspiration device is really necessary. Some have argued that it is faster to use an ultrasonic aspirator and I would counter that an osteotome and mallet would also speed up the resection; however, neither approach, in my opinion, is appropriately delicate enough for intraaxial brain surgery. The reason why conventional suction is generally safer, in my opinion, than ultrasonic aspirators, is that while both devices can and will put holes in arteries, I have found that conventional suction leaves holes which are smaller and easier to seal with bipolar spot welding techniques (described below) or other methods which preserve artery patency. I think this is because usually, the hole comes from the avulsion of a small side branching artery, which is manageable. Aspirator holes are often larger and harder to fix without occluding the artery (Fig. 2.2). Of course, with loupes, you may not really see any of these differences.

Method Two: Know the Vessel Anatomy and Work It into Your Planning

As stated above, you have to sacrifice some arteries to remove a glioma as you are removing a piece of the brain. However, it is always best to delay this sacrifice until you are certain of the vessels identity as a vessel supplying soon to be removed brain. You want to avoid sacrificing an artery that is supplying an important part of the brain beyond what you plan to remove. However, as you gain experience, it becomes easier to recognize which vessels pose some risk for being an en passage artery, meaning it is supplying brain which you do not plan to remove, and which ones almost certainly are not going to a different part of the brain. It speeds up the resection significantly when you do not fixate on the identity of anterior temporal MCA branches when the plan is for temporal lobectomy. It is more important when you realize you are dealing with an en passage situation ahead of time, and behave appropriately.

Common En Passage Situations (Fig. 2.3)

Midline frontal lobe: As the resection proceeds inwards towards the cingulate gyrus and corpus callosum, you begin to encounter more ACA branches, eventually reaching the ACA trunk. It is impossible the remove the corpus callosum, cingulate gyrus, or midline frontal lobe without addressing these branches safely. The majority of these branches are internal frontal branches, which are supplying the brain which you are going to remove; however, there are critical en passage branches, namely the callosomarginal artery, which are feeding the motor strip and SMA regions among others, which much be preserved. Here, it is critical to follow the rule that you should follow an artery to its target before taking it.

Frontal or parietal operculum: Removing an opercular tumor successfully is synonymous with safely working around the en passage arteries exiting the Sylvian fissure.

Insula: Accessing the insula requires working past the MCA branches which overlie it. Thus, most of the large arteries encountered are en passage arteries. Sacrificing the circumflex branches which supply the insula is critical for preserving the integrity of the numerous en passage vessels at risk in these cases.

Back of the Sylvian fissure: Dealing with tumors involving the supramarginal gyrus or superior temporal gyrus brings you to critical en passage arteries, including the branch I call the artery of death (described below).

Medial temporal lobe: Removing the uncus or parahippocampal gyrus puts arteries such as the anterior choroidal and PCA at risk, as they are supply blood to structures beyond the resection boundaries.

Arteries generally not en passage: 1) Vessels at the anterior temporal pole during temporal lobectomy, 2) Vessels at the anterior frontal pole during frontal lobectomy, 3) Vessels in the middle frontal gyrus (you are at the end artery by then), 4) Vessels at the occipitotemporal junction (you are at the end artery by then), 5) Vessels encountered in the medial parietal lobe (those vessels supply local brain), 6) PCA branches during an occipital lobectomy.

This list is not exhaustive and other situations can occur. The key is to always look at the T2 images ahead of time (Fig. 2.4) and get a sense of where the en passage arteries are in relation to the tumor. There are numerous ways to address these arteries and to get them out of the tumor, or at least not injure them. Removing brain in a subpial manner keeps a boundary which helps you identify the vessel and save it. Certainly, this is the standard method for complication avoidance in the medial temporal lobe. Additionally, while it is usually necessary to provide some counter traction to remove a tumor, knowing that an en passage artery is there probably should change your willingness to be aggressive in an area. In more difficult situations, such as high-grade glioma caked on the vessel, meningioma type surgical techniques, such as finding the vessel with a Doppler and serially dividing tumor next to the vessel may be necessary. In rare situations, it may be impossible to get the vessel out of the tumor with any technique.

Method 3: Avoid Manipulating High-Risk Arteries

Technically any artery can enter vasospasm, dissect, or tear when you manipulate it surgically; however, experience (my own and numerous other surgeons) has taught us that certain arteries are far more likely to cause disaster than others. In these cases, it is generally wise to leave these arteries alone, or at least to handle them with great care.

Lenticulostriate arteries (Fig. 2.5): These small branches of the proximal MCA supply critical structures such as the basal ganglia and internal capsule. Numerous surgeons have found that these arteries are particularly prone to cause a stroke when we try to remove tumor from them. Also, if you are encountering them in a glioma surgery, then there is a high chance you are in the basal ganglia, which seldom has a tumor in it, in my experience. As I emphasize at numerous points in this book, basal ganglia tissue looks and feels like glioma, and it is easy to continue into it unless you are always vigilant with considering the possibility that what looks like tumor may be basal ganglia. Thus, if you are seeing lenticulostriate arteries, you probably should rethink what you are doing there.

Anterior choroidal artery (Fig. 2.6): This tiny artery runs on the medial surface of the uncus and can undo an otherwise excellent resection of the temporal lobe by causing permanent injury to the internal capsule. Subpialization of the medial temporal structures is critical to avoid this, but if the pia is disrupted, then you should strongly consider stopping, especially with high grade gliomas and especially if the patient has previously had radiotherapy.

Optic nerve vasculature: This issue is seldom encountered in glioma surgery, but I have seen it occur a few times, and have seen surgeons get into trouble with optic nerve vascular injury in glioma cases. In short, the optic nerve has a tenuous blood supply, and loss of this blood supply can cause blindness. Avoid cautery when working on the posterior inferior frontal lobe, even when working subpially, as this can conduct through to the optic nerve and injury its blood supply.

“Artery of death” (Fig. 2.7): This is my name for the posterior temporal branch exiting the back of the Sylvian fissure and supplying the posterior temporal lobe. I have been haunted by this artery many times, notably by its ability to occlude or spasm right in front of me from even minimal manipulation. While it is always bad to occlude this artery, in speech mapping cases, particularly on the left side, the infarct can densely involve the language areas and end the mapping before you have really got a chance to get started. I have seen this infarct end in ultimately in good long term speech recovery, but presently do everything in my power to touch this artery as little as possible.