From the Temporal Side (Fig. 12.5)

A large fraction of insular gliomas co-inhabit the medial and/or anterior temporal lobes, and thus a large percentage of these cases will involve a temporal lobectomy. It is critical to note that I generally do not use the valuable time during mapping to complete the lobectomy. A patient will only map for so long, and you should use this time to focus where you really need to map. Plus, you will bump the dura multiple times taking out the lobe which causes pain, which in turn is distracting.

The initial phase of a temporal approach involves making a posterior temporal cut as described in Chapter 11. This involves first removing the STG from the posterior limit of the cut forward until you reach the anterior corner (i.e., the front of the Sylvian fissure) in front of the limen insula. This shows you where the insula is located, defines the vessels of the inferior Sylvian fissure, and helps locate the IFOF. In this phase, there are two key mistakes to avoid. First, the “artery of death” is usually leaving the Sylvian fissure and heading to the posterior temporal lobe, and this can completely undo a speech mapping case. Second, you want to keep an open mind regarding the short communications between a mapped site (especially language sites) and the IFOF or SLF as DTI tractography may not demonstrate these small fibers well and they can arc gently anterior or just dive directly medially. You should resect the STG as far posteriorly as the mapping will allow, as the posterior insula is both dangerous and easy to miss unless you make a concerted effort to find it. A fair bit of insula is tucked into the temporal junction.

At this point, you should extend the cut inferiorly across MTG towards the middle fossa floor. Once you know where the insula is, then finding the temporal horn is easy. This cut is complete once it extends into the ventricle or to the floor across the entire length of the cut.

Once the temporal lobe is out of the way, the next step is to elevate the frontal opercular cortices. I let the patient rest during this period. I do this by gently retracting the opercula off of the insula bluntly, i.e., by padding the cortex then sweeping to pull apart the arachnoid of the Sylvian fissure. You should be careful to not over retract across the plane of an artery leaving the fissure; however, with care this maneuver is usually safe and effective. You should continue doing this from anterior to posterior until you clearly see the circular sulcus on all sides. If you have not seen the sulcus and its groove where the opercular cortices turn back on themselves to overlie the insula, then there is a very high chance you will miss a fair bit of the insula. Of particular importance, is the need to see the posterior superior and anterior superior portions of the circular sulcus, which are much higher underneath the frontal opercula than you think. If you cannot see them then you will either miss a great deal of tumor, or you will be blindly sweeping near the SLF or corticospinal tract, both of which are suboptimal ways to operate in the insula.

You should be aware that approaching from the temporal side, your attack angle is running slightly parallel to the cortical surface, with a medial angle, which you should take into your calculations as you descend into the insula (Fig. 12.6). The angle is more medial inferiorly and more flat as you progress superiorly.

From the Frontal Side

This is the less desirable approach, as it requires you to tangle with potential speech and motor sites, which adds to the risk. However, if those opercula are involved and are spilling into the insula, then there is no way to adequately address these tumors without a frontal approach. Also tumors which a large superior extent are often very difficult to tackle from the temporal side alone, and sometimes adding a frontal approach is the only way to get a good exposure.

The frontal approach (Fig. 12.7) involves slow, subpialization of the opercula as dictated by the mapping and tumor anatomy. Usually, if a cortex has mapped negative during the define phase, then it can be resected down to the level of the insula. It is also acceptable to resect portions of the subcentral and lower sensorimotor cortices if absolutely needed (they are bilaterally represented and well tolerated). There are numerous arteries intertwined with these cortices, and many of them are going to the motor cortex, speech, motor planning, and other critical sites.

Uninvolved opercula (i.e., ones which are being removed for exposure but not for tumor resection) do not need to be gross totally resected, but instead need to be sequentially trimmed from bottom to top until you can see their pial bank curving into the circular sulcus. A debulked sulcus is often enough to mobilize the pia and arteries and get them out of the way.

An analogous sweeping of the STG off of the insula is then performed with the caveat that the “artery of death” may limit how much you can do this, and you may need to remove some STG as well to get enough room.

Note that while the attack angle of the frontal approach also parallels the insular cortex, it is more consistently parallel than the temporal approach (Fig. 12.8). In other words, you are often looking at the side of the insula, instead of directly medial to lateral as your mind thinks of it.

From Both Sides

Obviously, there are numerous possible permutations depending on the size and anatomy of the tumor and the functional brain anatomy. In most combined cases, I do the temporal side first, because it makes a lot of room quickly, and this shows me where the insula is which assists with the other portions.