3 Surface and Gross Anatomy of the Cerebrum
The contents of this chapter have been extensively covered by numerous authors, including Rhoton. This chapter is included not to redo this work, but to provide a reference, to point out a few features of sulcal, gyral, and structural anatomy which is not usually covered in other work, but which I think is helpful to glioma surgery, and most importantly to demonstrate how to practically apply this anatomy to precisely remove these tumors. In general, I have tried to make this section shorter and more practically focused than other similar chapters; however, I highly recommend supplementing this reading with more fundamental texts if the content of this brief overview is not entirely familiar to you.
3.2 Why Study Gross Cerebral Anatomy in the Image Guidance Era?
Besides the obvious possibility that the image guidance might be inaccurate, and the fact that the tool is only as useful as the brain interpreting it, it is critical to have a sense of what anatomy exactly needs to be removed, and what potential functional consequences there are for doing so. By infiltrating and expanding brain structures, gliomas can radically alter the appearance and organization of the brain. They often distort the anatomy to such an extent as to completely obscure their origin and exact anatomy at first glance. For example, it is not at all rare to see a glioma which at first glance appears to be a motor region tumor, but in fact is a parietal glioma, which has so distorted and expanded, the anatomy that it has pushed its way into the expected location of the motor strip. A good example of such a situation is the sizeable GBM in Fig. 3.1. A quick glance at the contrasted images in Fig. 3.1 might lead you to conclude that this tumor is beginning in the back of the temporal lobe and is invading the basal ganglia and internal capsule. However, looking at the T2 images (which are better at defining sulcal anatomy), this tumor seems to be arising from the angular gyrus which is massively expanded. The precentral and postcentral gyri are both pushed forward and not heavily involved. Most critically, the basal ganglia and internal capsule are also pushed anteriorly (as we would expect an enlarged angular gyrus to do). Based on this, we were able to get an excellent resection of this tumor leaving the patient only with the homonymous hemianopsia he entered with. As should be obvious, careful observation of the preoperative images can radically alter your expectations.
Additionally, as I repeatedly point out in this book, the functional organization of the cerebral cortex is not entirely predictable based on the anatomy, but at the same time, it is not random. Even more than inter-individual variability, the presence of a glioma, especially oligodendroglioma but any glioma, can cause the functional anatomy of the brain to reorganize sometimes drastically; however, you should not expect to find the primary motor function at the occipital pole. Thus, while we need to clear our mind of the poison of exact anatomic expectations in the neocortex, knowing gyri and sulci gives us a reasonable framework to start with.
Finally, in a paradigm like this, where resection continues until encountering an anatomic or functional boundary, sulci are godsends. They provide an orienting landmark, and in some cases following a sulci can help keep you from crossing into the brain where you do not wish to be. Sulci show you where unseen vessels may be lie, give you a boundary to perform subpial dissection, and as long as you remain above the bottom of the sulci, you generally are not into the deep white matter tracts. Similarly, knowing the relationship of the brain to the ventricles, the basal ganglia, and other critical deep structures is essential for keeping out of trouble.
The bottom line is that the cerebrum does not have a lot of landmarks, and you need to be prepared to use whatever landmarks you have available.
3.3 Anatomy of Cerebral Sulci and Gyri
When providing a list of gyri and sulci and some descriptions, it’s best to start with some general observations and disclaimers. First, the term gyrus is best viewed as a concept which describes the basic direction and anatomic pattern of a strip of cortex that is usually identifiable in most brains. Thus, some gyri are more consistent and uniformly coherent than others, and to not understand the limits of the term is to try to force uniformity on anatomy which is frequently quite variable. For example, most people mapping the motor strip for the first time are surprised to find that motor function seldom maps to a single continuous strip of the brain, as the precentral gyrus is often interrupted by variable numbers of small sulci.
Another critical general observation is that almost all sulci point at a ventricle. The only exception to this is the sulci outlining limbic cortices of the medial hemispheric surface which instead parallel the corpus callosum. This is especially important when we begin to discuss the division phase of subcortical mapping in future chapters. In this phase, our plan is to continue mapping the patient until the cut is completely past the white matter tract at risk, and thus the tumor specimen has been made non-eloquent. Entering the ventricle is often the way to assure that you are past it as there is no way the tract runs through the ventricle. In many such cases, the true danger occurs once the dissection proceeds beyond the depth of the sulci, as the large fiber tracts are there, and there are often no landmarks. While one guide to preventing getting lost in the brain is functional mapping, keeping your cut parallel to the sulci is another guide to quickly getting into the oriented safety of the ventricle.
You might be surprised when I say that it is a good idea to enter the ventricle, as we are all taught not to do this. Certainly, this maneuver increases the risks of needing a shunt, and you should not do this for no reason; however, it is easier to place a shunt than to fix an incomplete resection, and it is much easier to place a shunt, than to rehabilitate someone whose brain you got lost in because you didn’t want to use the ventricular anatomy as a landmark.
What follows then are some basic descriptions of the gyral and sulcal patterns of various lobes, with general rules given whenever possible.
3.4 Surface Anatomy of the Frontal Lobe
The human frontal lobe is massive, as anyone who has resected a frontal lobe likely has observed. Its lateral surface is best known and is dominated by three large anterior to posterior gyri (superior, middle and inferior frontal gyri) which all about the precentral gyrus in a perpendicular manner on their posterior edge. These are divided by the superior and inferior frontal sulci.
When attempting to orient myself to the anatomy of any image in this region (Fig. 3.2), I always find the superior frontal sulcus first, as this knowledge is critical for determining where the precentral gyrus (i.e., the motor strip) is, where the supplementary motor area is likely to be, and for figuring out which frontal gyrus is which. Numerous other gyri can be sorted out using deduction based on this fact.
The superior frontal sulcus is deep, usually continuous, and intersects the precentral sulcus at a right angle (which is what makes it so helpful to figuring what is what). Its continuation into the pre-central gyrus indents that gyrus in what’s known as the hand knob (but which cannot be blindly considered to be hand motor cortex, especially in glioma patients).
In contrast, the inferior frontal sulcus is always interrupted, and always partially indents the inferior frontal gyrus to form the triangularis portion of that gyrus. The junction of the inferior frontal sulcus with the precentral gyrus provides another cortical landmark. Inferior to this point, the precentral gyrus in normal people is usually face motor cortex (which can be sacrificed in many cases), and usually is operculum which overlies the insula. The inferior frontal gyrus has its three well known divisions: pars orbitalis, pars triangularis and pars opercularis.
The medial surface of the frontal lobe (Fig. 3.3) is dominated by the medial border of the superior frontal gyrus which extends from the motor regions around to the orbital floor in a c-shaped fashion. The cingulate sulcus (one of the most useful sulcal landmarks in my opinion) separates the cingulate gyrus from the superior frontal gyrus. The cingulate gyrus parallels the corpus callosum and is separated from it by the callosal sulcus. Once it continues under the rostrum of the corpus callosum, there are three small gyri which overlie structures of the basal forebrain such as the septal nuclei and nucleus basalis.
The orbitofrontal surface of the frontal lobe is generally delineated by an H-shaped sulcus that divides the orbitofrontal cortex into four. The olfactory tract separates these gyri from the gyrus rectus which is medial to the tract. The posterior aspects of these cortices blend into the anterior perforated substance in the proximity of the optic nerve, and this is the location of the basal forebrain, and caudate nucleus. I have not spent much time describing these gyri in much detail because personally, I have not found them to be surgically very useful, and I think of the orbitofrontal cortex as a unit.
3.4.1 Functional Considerations
Care should be taken to not firmly link specific places in the cortex with specific functions; however, in this chapter I will make some basic comments on the specific location of various functions, and a more nuanced view will follow in subsequent chapters (Fig. 3.4).
The frontal lobe has been considered mostly non-eloquent by neurosurgeons for some time; most of whom I suspect has not spent a great deal of time talking to family members of patients with frontal lobe injuries. The frontal lobe is highly eloquent, it is just highly complex, the problems its loss causes are less obvious that other parts of the brain, its regions likely have some degree of redundancy, and we understand very little about how it works or how to study it.
Textbooks on frontal lobe function divide the frontal lobe into three basic areas: Motor (the motor strip), Premotor (areas just in frontal of the motor strip involved in motor planning), and Prefrontal (all the thinking, motivation, judgment and emotional regulation parts in front of premotor areas), and for our purposes this is useful.
Prefrontal cortices can be further subdivided by functions conventionally attributed to them. The dorsolateral prefrontal cortex (DLPC) (i.e., the lateral convexity) has a complex relationship with working memory and concentration that is not well understood. The exact location the regions which make up the DLPC proper is hard to nail down from fMRI data; however, the posterior middle frontal gyrus just in front of the premotor areas is a common place this function localizes to. The medial structures, the medial superior frontal gyrus and especially the cingulate gyrus, contain brain networks involved in motivation and attention. The orbitofrontal cortices are involved in emotion and emotional regulation (i.e., the problems we see after bifrontal contusions which are classically an orbitofrontal injury).
Premotor cortices constitute a complex network which we will discuss in more detail in subsequent chapters. The supplementary motor area classically is located in the medial bank superior frontal gyrus just in front of the precentral gyrus. Its general vicinity can be estimated by locating the part of the superior frontal gyrus which joins with the precentral gyrus. The other premotor areas are located laterally just anterior to the motor strip, in the posterior edges of the middle and inferior frontal gyri, and this network has a highly variable anatomy.
Broca’s area is classically defined as the inferior frontal gyrus just anterior to the motor strip. To give this site any more significance other than it is the premotor area which plans movement for the face, tongue, palate, and larynx is to misunderstand the role of this site in the speech network. As with any other cortical site, the site which causes speech arrest with cortical stimulation can reorganize and move to other parts of the brain, and thus it is sometimes located in the canonical Broca’s are, but it can be many other places. One thing that is usually true is that it is the premotor area for the face, and usually finding facial movements will help you locate the speech arrest site which is generally anterior to the face motor area (logical, no?).
3.4.2 Important Distinctions in Frontal Glioma Anatomy
In every area of the brain, distortions of sulcal anatomy can make it challenging to determine where a tumor is really centered and what it really involves. Sometimes, this is academic, and other times it is easy. In each section of this chapter, there are examples of situations where the distinction is often challenging and the difference between two gyri make a big deal.
The most common scenario are tumors of the posterior frontal lobe where the decision needs to be made whether this tumor involves the motor strip, sensory strip, premotor areas, etc. As described above, the key is locate the superior frontal sulcus and use this to define the precentral sulcus, which may be smashed (Fig. 3.5 , Fig. 3.6) and may be displaced remarkably posterior. Knowing whether you are dealing with a motor strip or premotor tumor can make a significant difference in your mapping strategy and plan.
The other common situation with frontal gliomas is differentiating whether the tumor arises from the superior frontal gyrus, the cingulate gyrus, or both. Experience has taught me a great deal of respect for the cingulate gyrus, especially if it is uninvolved with glioma, and I spend a fair bit of time sorting out this distinction. The exact status of the corpus callosum should also be studied, as many cingulate tumors will herniate under the falx and mimic a butterfly tumor when they really aren’t. The last thing you want to do is to drift into the contralateral cingulate gyrus thinking somehow they are contiguous; this is easy to do when the tumor is bloody and the contralateral pia is abnormal. The key is to study the coronal images closely, as the axials can be confusing especially in giant medial frontal tumors. Three examples are provided in Fig. 3.7 , Fig. 3.8 , Fig. 3.9.