The general technique of aTLY has been noted elsewhere (Girvin 1992). With some modifications, it will be similarly repeated here. The modifications will include some alterations of surgical anatomy, through both the text and figures.

There are a number of different scalp flaps used for access to the temporal lobe. I will only note that which I believe to be the best. It is the one that I learned at the Montreal Neurological Institute and is demonstrated in Fig. 6.1. It is certainly the best flap from the cosmetic point of view and thus it has significant advantages, particularly in the case of young women. It lacks having to cut through the temporalis muscle, which is required to some extent in any T-shaped incision. It also precludes in the latter the potential of a vertical anterior arm of the flap coursing just inside the hairline anteriorly. Thus, I do feel that the described flap provides a better cosmetic result. In addition, I believe that the larger single flap is easier to retract and to close at the termination of the operative procedure.

To reiterate, the proposed scalp incision is injected with 0.33 % bupivacaine hydrochloride (Sect. 3.​3.​4). When the incision is made, bleeding arteries can be coagulated and the use of peroxide gauze, with rare exception, is all that is needed for the remainder of hemostasis if the initial scalp injection has been satisfactory (see Sect. 3.​6.​3). (If the surgery is to be done under local anesthesia, see Sects. 3.​3.​2, 3.​3.​4, and 3.​6.​2.)

The craniotomy, using local anesthesia and neuroleptanalgesia, has been outlined earlier (Sect. 3.​6). The sphenoid wing of the temporal bone creates a special difference between the routine craniotomy on the one hand and the ideal craniotomy for the removal of an anterior temporal lobe. I had initially performed the craniotomy in what I believed was the usual fashion, placing a burr hole on either side of the sphenoid wing and then simply running the Gigli saw, or craniotome, between the two holes. When the bone flap was removed the final part of the craniotomy was composed of a large removal of the outer portion of the sphenoid wing and the inferior temporal bone with rongeurs. This was a perfectly reasonable method for such a craniotomy. However, I came to be very often disappointed in the final cosmetic appearance of the characteristic depression in the skin over the anterior temporal region. In the case of young girls, my disappointment was the greatest and it soon became associated with embarrassment. I eventfully considered that this should not have to be the case!

Given the fact that I performed nearly all the temporal craniotomies under local anesthesia and the fact that under the latter the inferior dura is the most important dura to be anesthetized for the relief of discomfort, I began using the air drill for making the inferior bony incision. The normal burr holes are placed inferiorly and in the so-called “key hole” area above the base of the sphenoid wing. Using the medium-sized diamond burr in the air powered drill and small rongeurs, the inferior bony incision line is achieved in a series of steps, involving the intermittent injection of local anesthetic between the leaflets of the dura. As noted earlier (Sect. 3.​6.​7), it is very advantageous to initiate the craniotomy inferiorly in order to anesthetize the inferior dura in a temporal craniotomy, as this usually is nearly all the dural anesthesia that is required for the craniotomy. (If the craniotomy is initiated in the superior part, then local anesthesia will be required two times, e.g., superiorly and then later in the inferior part.) The line is taken anteroinferiorly through whatever part of the sphenoid wing that needs removing; a large part of the sphenoid wing is included in the bone flap (The resulting craniotomy, from this simple small linear incision through the sphenoid wing, never results in any unwanted cosmetic suggestion postoperatively of an individual having had a craniotomy.).

One other feature that I have found to save significant time of operation, if using sutures for closing the bone, is to, just prior to raising the bone flap, mark the holes for the sutures and drill them after raising the flap, something that takes no more than about 30 s. (It is one of the reasons that allow a satisfactory closure in ~30 min.)

I open the dura in a simple X shape, with the bases of the X being in the corners of the craniotomy, fold them, and use rubber bands for retraction. I have used other openings, but I have found this to provide the easiest satisfactory closure for me.

Walker (1967) advocated making the initial incision through the middle temporal gyrus (MTG), thus preserving the superior temporal gyrus (STG), but gave no reason for doing so. Some have adopted this, often giving the reason as that of “safety,” e.g., decreasing the likelihood of injuring the underlying middle cerebral artery vasculature (MCA). I have seen exactly the opposite on more than one occasion, when the middle temporal gyral path had been utilized, in order to save the STG! In two or three of these instances, the incision was taken too inferiorly, as it was directed medially, and ended up below the lower edge of the leptomeningeal investment of the medial surface of the superior temporal lobe, in among the insular cortex, external capsule, and on one occasion the internal capsule. The ease of this eventuality can be appreciated from viewing Fig. 6.2, an illustration of a coronal view through the middle of the temporal lobe; a perpendicular incision through the MTG could quite easily end up inferior to the bottom of the vertical limb of the Sylvian fissure. Quite apart from the fact that I believe this to be a dangerous approach, especially in the hands of a less experienced surgeon, it also leaves tissue (primarily the STG) that is at least partially denervated, i.e., de-afferented and/or de-efferented—a potential for the origin of epileptogenicity and thus a contraindication to the principles of good epilepsy surgery. There is no good reason for saving the STG! Thus, I dogmatically believe that the initiation should be through the STG for reasons discussed in the following, especially now with the current practice of only very rarely extending the posterior resection line in the STG beyond ~4 cm posterior to the temporal pole.

Using an initial incision through the STG allows the surgeon to expose the leptomeninges not only over the lateral surface but also over the superior and medial surfaces, i.e., the horizontal and vertical limbs, respectively, of the Sylvian fissure, right from the beginning of the procedure. Therefore, the surgeon can take full advantage of the anatomy and the protective value of the leptomeningeal barrier by remaining within the confines of the leptomeninges of the horizontal and vertical parts of the Sylvian fissure. Through the leptomeninges the underlying anatomical structures, e.g., MCA branches and insula, can always be visualized. Further, it also provides ready visualization of the boundaries of leptomeningeal investment of the Sylvian fissure with the progression of surgery.

Often the superficial middle cerebral vein is large, follows the Sylvian fissure, and empties into the sphenoparietal sinus anteriorly or into draining veins posteriorly, such as the vein of Labbe. It important to be aware that it may also veer off the Sylvian fissure, particularly more posteriorly, and lead to misinterpretation of its course in conducting an aTLY or inferior Rolandic corticectomy, i.e., some of the superior temporal gyrus may appear above the vein or a part of the frontoparietal operculum below the vein! The best accurate strategy under this circumstance is to follow the leptomeningeal barrier, which has been incised more anteriorly, as it defines the Sylvian fissure, irrespective of the course of the superficial middle cerebral vein.

The leptomeningeal investment of the lateral aspect of the STG is coagulated in an anteroposterior direction about 3–5 mm below the Sylvian fissure, within the proposed AP extent of removal (Fig. 6.3). Small incisions are then made in the coagulated leptomeninges in between the MCA branches, which emerge out of the Sylvian fissure over the STG and which supply a part of the temporal lobe parenchyma being removed. There are usually three or four such vessels. Through these incisions the subpial dissection of the underlying STG is initiated. Removal of parenchyma from beneath the arterial branches overlying the STG isolates them, and 3–5 mm provides adequate length for their being safely picked up with their overlying leptomeninges, coagulated, and incised. The incision throughout the extent of proposed STG removal is now relatively complete. This now provides a clear, unimpeded path for the advancement of the subpial dissection.

Using the barrier provided by the leptomeninges, the subpial dissection initially is taken over the superior surface, thus separating the STG from the horizontal limb of the Sylvian fissure and the frontal operculum (and the anterior parietal operculum in more radical aTLYs, e.g., posterior to the Rolandic fissure). The continuation of dissection then separates the medial surface of the STG from the leptomeninges of the vertical limb of the fissure, thus separating the medial aspect of the STG from the underlying insula and the MCA vessels. In the whole of this resection, the double layer of leptomeninges is usually easily preserved. This layer, which had covered the STG, serves to provide a sturdy protective covering of the MCA vessels and the insula. As noted in Sect. 2.​2.​3, this may be achieved by blunt dissection or by suction or perhaps most commonly by a combination of both. This is illustrated in Fig. 6.4, which is very similar to the illustration used earlier in Fig. 2.​6. The subpial dissection, removing first the superior and then the medial STG cortex, is carried inferiorly until the symbol ✱ is reached, which is illustrated in cross sections in Figs. 6.2 and 6.4. This point represents three important anatomical features: (1) the inferior aspect of the leptomeningeal investment of the medial temporal (STG) cortex, (2) the bottom of the vertical limb of the Sylvian fissure, and (3) the superior aspect of the temporal stem. Interestingly, this point is nearly always only 1 mm or so below a large, primarily anteroposteriorly oriented, MCA vessel within the Sylvian fissure. During the dissection this point is easily recognized by the fact that there ceases to be evidence of the leptomeningeal barrier combined with the uncovered appearance of the very white temporal stem (see also Fig. 6.5b), which is composed of the connecting fibers of the anterior part of the temporal lobe to the interior of the hemisphere, e.g., orbitofrontal cortex, hypothalamus, anterior commissure, inferior occipitofrontal and longitudinal fasciculi, etc. Also, at this juncture there is often some very minor venous bleeding, which can be simply packed for a minute or so with a cottonoid patty; rarely is coagulation necessary. However, since this is also the line along which the temporal stem will be incised further on, it is often advantageous to simply coagulate a shallow (~1–2 mm) incision through the stem at this point, along the bottom of the leptomeninges. The best direction for deepening this incision can be better estimated later in the resection.

Once the vessels overlying the STG are coagulated and incised a large majority of the blood supply to the anterior TL will have been interrupted. In order to isolate and devascularize the majority of the remaining part of the most anterior temporal lobe, the leptomeningeal incision of the STG is now carried anteriorly by subpial dissection associated with coagulation of vessels over it. Providing that the leptomeninges are incised as far as the dissection has progressed, then at this point inferolateral retraction of the anterior aspect of the lobe from beneath the sphenoid wing facilitates better exposure of the operative field. With the continual anterior extension of the subpial resection of the STG along the underside of the sphenoid wing, a point is reached where there is no longer any juxtaposed frontal operculum on the other side of the leptomeninges, but rather simply the anteromedial bony wall of the anterior middle fossa, covered with the dura mater. Continuing further, eventually the floor of the middle fossa is reached. At this point the free (isolated) leptomeningeal membrane can now be gradually incised coincident with the progression of the subpial resection. (If the patient is under local anesthesia and coagulation is required, the membrane must be lifted away from the dura for the coagulation, or the dura anesthetized, so that it does not result in discomfort.) Eventually, with this progression, the dissection will extend anteriorly, then leave the Sylvian fissure inferiorly around the anterior part of the temporal stem, and finally move somewhat posteriorly. This is illustrated in Fig. 6.6a, the point at which the part of the TL anterior to the temporal stem is partially freed as a result of the subpial dissection and incision of its leptomeningeal investment medially. With retraction of the anterior TL at this point, not only can the superolateral aspect of the temporal stem be seen at the bottom of the anterior Sylvian fissure, which was initially seen when the STG was being resected (Figs. 6.4 and 6.5), but now its anterior extent can be visualized. Thus, the subpial resection can be taken around the anterior extent of the temporal stem, which actually occurs between Fig. 6.6a, b. Continuation of the dissection will now progress posteriorly, under the stem, back along the underside of the inferomedial surface of the TL.

At this juncture, approximately 2 cm behind the temporal pole, the subpial dissection will be continued posteriorly over the inferomesial aspect of the temporal lobe separating the parenchyma from its leptomeningeal investment. Continued posterior dissection at this point is over the bony floor of the middle fossa and eventually the anteromedial tentorium cerebelli. This provides some increased exposure along the deep inferomedial surface of the anterior temporal lobe—one of the most important parts of the surgery of the aTLY, but the surgeon cannot easily visualize the inferomesial structures being stripped of their leptomeningeal investments. Taken about as far as the dissection can be carried out at this point, the vicinity of the uncus will have been reached. It is difficult to continue the dissection further posteriorly because of the bulk of the temporal stem (see also Fig. 6.6b).

If the aTLY is being carried out under local anesthesia, nearly always once there is retraction required on the anterior part of the lobe, when rounding the anterior end of the Sylvian fissure, or carrying out the subpial dissection of the inferomedial side of the lobe, the patient will begin to experience discomfort. It is at this point that the injection of some local anesthesia into the anterior part of the tentorium, lateral to the third nerve (which can be seen through the leptomeninges), into the area of the trigeminal nerve will be somewhat painful, but it is usually the last requirement for the use of local anesthesia in the operation of an aTLY. This is shown in Fig. 6.7, which illustrates the cranial base. The very slow injection of 0.5–1.0 ml of anesthetic can be administered through the use of a long #25 needle into the region, between the leaflets of the dura, within the dashed circle of the figure. This is very close to the underlying branches of the ophthalmic division of the trigeminal nerve, emerging in this area, that give rise to the most important innervation of the tentorium, as shown in Fig. 3.​4. As outlined by Feindel and his colleagues it “forms a major contribution to innervation of intracranial structures” (1960, p. 563). It is likely the fact that the needle is inserted into this bundle of nerves, within the dura leaflets, which is usually somewhat painful. Because of the likely discomfort of the injection, I always warn the patient at the time of this likelihood. However, the benefit of the injection far outweighs the very brief discomfort, as it nearly always abolishes not only the immediate discomfort of the patient but is often sufficient to abolish any residual discomfort associated with the finishing of the operation. Infrequently there will be a partial anesthesia of the trigeminal, oculomotor, or trochlear cranial nerves, which are in the very near area of the injection, which may last for a few hours.

At this juncture the anterior aspect of the free edge of tentorium, probably the third cranial nerve and perhaps the internal carotid artery, can be seen through the partially translucent leptomeninges from which the cortical parenchyma has been removed. The farther posterior that this subpial resection is able to be conducted on this (deep) underside of the medial temporal cortex, the easier will be the later part of the aTLY dealing with the incision through the bulk of the temporal stem and then eventually the removal of the anteroinferomesial (limbic) parenchyma. The ease of visualizing these structures and the reduction in the necessity of excessive retraction of the most anterior part of the neocortex can be facilitated by an ongoing associated coagulation and incision of the leptomeninges, which have been freed and isolated by the subpial resection. The incision at this point should be conducted roughly parallel to the free edge of the tentorium and a few millimeters lateral to it, i.e., thus leaving its free (incised) edge covering the free edge of the tentorium.

The intact temporal stem eventually creates difficulty of continuing further posteriorly along the underside of the inferomedial structures. This can be at least partially rectified by initiating the incision through the anterior aspect of the temporal stem and gradually extending it posteriorly to the posterior extent of the subpial incision, ending in the vicinity of the pebbled line (•••••) depicted in Fig. 6.6b; that is to say, the TS incision line can be safely conducted from the bottom of the Sylvian fissure through the stem to the furthest posterior point of the dissection over the inferomedial surface. This incision through the anterior stem will form the initial incision the completion of which (vide infra) will isolate the amygdala from the anterior temporal lobe. The surgeon will not be aware of the presence of the amygdala at this point, but in fact it is juxtaposed to the anterior part of the uncus and actually accounts for part of the medial protuberance of the uncus. However, the fact of the amygdala at this time is not something that can be appreciated by the surgeon. After the incision the uncus/amygdala will have been separated from its medial leptomeningeal investment. Its final isolation at its base will be achieved after the removal of the neocortex (vide infra). Thus, while the incision of the anterior stem and subcortical white matter must be “tight” to the superficial (above) and deep (below) leptomeningeal barriers, respectively, yet it must not encroach on the surrounding parenchyma. At this point, an alternative to continuing the dissection more posteriorly along the inferomedial surface, one may choose to put a cottonoid patty in the resection bed of the incision in the anterior stem and return to the posterior resection extent of the STG and proceed to establish the posterior resection line (PRL).

In my view the preferred alternative is to continue with the more posterior dissection along the deep part of the medial surface, only if it can be conducted satisfactorily. If this alternative is chosen, then the adequacy of progressively viewing this dissection is not only facilitated by, but also dependent upon, the coincident continuation of the incision through some of the temporal stem subcortical white matter between the two boundaries noted in the foregoing paragraph. With very little more posterior dissection, the choroidal point will be reached, which is just outside the medial aspect of the ventricle at the posterior part of the amygdala. This is illustrated in Fig. 6.6c. Although perhaps the surgeon will be unaware, a coronal section at this point would reveal that the incision is posterior to the protrusion of the amygdala into the lateral wall of the ventricle and posterior to the anterior part of the pes (head) hippocampus. Further, at this point the lateral surface of the brain stem (midbrain) may be visualized through the double layer of partially translucent leptomeningeal barriers, e.g., that adherent to the lateral side of the midbrain peduncle and that which has been separated from the medial side of the TL. (This is also a location where the optic tract, coursing anteroposteriorly from the optic chiasm to the lateral geniculate body, is less than a millimeter away!) If the incision through the temporal stem and basal amygdala is sufficiently posterior, the “mantle” of the ventricle at this point is less than 1 mm in thickness, as seen in Fig. 6.6c. The “mantle” here is the last tissue that needs to be incised (into the ventricle) in order to complete the deep anteroposterior separation of amygdala and uncus from its overlying leptomeningeal investment. The entry of the ventricle through this thin mantle now wholly isolates the anterior temporal lobe anteriorly. Its only neuronal connectivity is that now coursing through its posterior parenchyma. At this point, the most difficult part of the aTLY has been achieved!