Number of patients
23
Mean age
49 years (range: 25–75)
Male: female
15: 8
Presenting Karnofsky Performance Score
80
Presenting symptoms
Speech deficit
5
Visual deficit
2
Seizure
15
Table 2
Tumor characteristics and approach
Right: left lateralization | 10: 13 | |
Mean size (CC × AP × ML) | 2.9 × 3.5 × 2.8 cm | |
Cystic | 6 (26 %) | |
Necrosis | 12 (52 %) | |
Schramm distribution | ||
A | Tumor limited to medial temporal lobe; extends from uncus to hippocampus/parahippocampal gyrus | 13 (57 %) |
B | Considered to be paramediobasal tumors – they lie lateral to type A tumors and the collateral sulcus | 2 (8 %) |
C | These tumors occupy the same anatomic regions as type A and B tumors | 8 (35 %) |
Site of corticectomy | ||
Middle temporal gyrus | 9 (39 %) | |
Inferior temporal gyrus | 14 (61 %) | |
Tumor histology | ||
WHO grade III | 14 | |
WHO grade IV CC: Carniocaudal AP: Anteroposterior ML: Mediolateral | 9 |
Surgical Approach and Outcomes
The inferior temporal gyrus was the most commonly employed site of cortical entry for tumors in the cohort (Table 2). Gross total resection was achieved in 92 % of the population (Table 3). With regard to complications, the most common events were: clinically significant stroke (two patients), new visual deficit (two patients), and new speech deficit (one patient). One patient without a prior history of seizures suffered a postoperative ictal event; seizures were controlled in only 53 % of those patients suffering preoperatively with seizures. The relationship between complications and site of cortical entry was further assessed (Table 3); a majority of events were encountered via the middle temporal gyrus approach. Statistical analysis revealed p values of 0.14 and 0.39 for differences in visual outcomes and speech outcomes between middle temporal gyrus and inferior temporal gyrus cortical entry.
Table 3
Surgical outcomes and complications
Outcome | # (%) | ||
Extent of resection | |||
Subtotal resection | 2 (8) | ||
Gross resection | 21 (92) | ||
Seizure control rate | 8/15 (53) | ||
Complications by approach | Middle temporal gyrus | Inferior temporal gyrus | |
New postoperative stroke | 2 (8) | 1 | 1 |
New speech deficit | 1 (4) | 1 | 0 (p = 0.39) |
New visual deficit | 2 (8) | 2 | 0 (p = 0.14) |
Discussion
The MTL represents a unique structural entity within the supratentorial space with regard to the profile of brain tumors affecting this region. Due to its relatively compact structure and the fact that it is surrounded by critical vascular structures and white fiber pathways, much consideration has been paid to identifying ideal surgical approaches to the region with minimal collateral damage [1, 6, 8, 13–16, 21, 22, 26–28]. In an effort to better understand the risks of transcortical approaches through the middle/inferior temporal gyrus for high-grade gliomas, we reviewed our institutional experience with regard to the following outcomes: extent of resection, postoperative visual field deficits, and postoperative speech deficits.
White Fiber Tracts
Important white fiber tracts surround the MTL and it is important that they should be considered when selecting a surgical approach. These tracts include the visual pathways, uncinate fasciculus, superior longitudinal fasciculus, and inferior longitudinal fasciculus. While the location of Wernicke’s area has been well documented and can be determined intraoperatively through awake mapping, considerable attention has been paid to elucidating the location of white fiber tracts that cannot be found via intraoperative mapping.
Considered a part of the limbic system and temporal stem, the uncinate fasciculus, which lies in close relation to the amygdala and limen insulae, is an often a neglected entity in temporal lobe surgery. This white fiber tract consists of three parts: a ventral (frontal) extension, an intermediary segment within the limen insulae, and a temporal segment [18]. This fiber tract is asymmetrical in size between both hemispheres (27 % larger in the right hemisphere), suggesting its importance even in non-dominant hemispheres [18]. Interconnecting the anterotemporal lobe with the orbitofrontal area, the uncinate fasciculus is involved in linking emotions to cognition [6, 7, 20]. It may also play a role in the retrieval of episodic memories [25], and surgical resection of lesions in this area can result in long-term difficulty with face recognition and object naming [17]. Disruption of the uncinate fasciculus with anteromedial temporal lobectomy or trans-sylvian transinsular approaches may be associated with the psychosocial and cognitive changes seen postoperatively [1, 2]. As part of the temporal stem with its origin in the hippocampal formation and the amygdala, the uncinate fasciculus may be the preferential pathway for seizure spread. This must be recognized when the goal of surgery is to control tumor-induced seizures [12].
Superior and superficial to the uncinate fasciculus, the inferior occipitofrontal fasciculus (IFOF) is believed to be involved in semantic processing [11]. The IFOF consists of two components – a dorsal subcomponent connecting the frontal, superior parietal lobe and occipital gyri, and a deeper ventral portion communicating with the visual association areas: inferior occipital gyrus and posterior temporal-basal area (fusiform gyrus, temporo-occipital sulcus, and basal inferior temporal gyrus). Intraoperative stimulation studies have shown that IFOF stimulation induces semantic paraphasias during picture-naming tasks [3, 4].
Advances in diffusion tensor imaging have further elucidated the anatomic projections of the inferior longitudinal fasciculus (ILF), which joins the posterior occipital temporal regions with the temporal lobe, where it further interacts with the uncinate fasciculus to interact with the basal frontal region. Within the dominant hemisphere, the ILF is proposed to be one of the parallel pathways of the “semantic ventral stream” that constitutes the language circuitry. Intraoperative mapping studies by Mandonnet et al. elicited semantic paraphasias when white fiber tracts underneath the superior temporal sulcus immediately above the roof of the temporal horn were stimulated [10].
Numerous recent studies utilizing Klingler dissection and diffusion tractography have helped to elucidate the course of the optic radiations as they project from the lateral geniculate body (LGB) to the calcarine cortex [6, 24]. As they leave the thalamus, the fibers can be classified into three bundles: posterior, central, and anterior – which progressively take a more curved route to the calcarine cortex. The central group of fibers follows a direct path from the LGB to the visual cortex without any anterior curve, while the central bundle takes a partial anterior curve over the superior surface of the temporal horn (just at the level of the LGB) prior to extending posteriorly. The bundle of most concern during surgical approaches to the MTL is the anterior bundle (Meyer’s Loop) that passes entirely around the lateral half of the tip of the temporal horn before coursing posteriorly within the sagittal stratum to the calcarine cortex. Debate exists with regard to the precise relationship of the anterior bundle to the tip of the temporal horn [2, 5, 9, 19, 24]. Some authors have reported that Meyer’s loop does not reach the anterior tip, while other studies have noted that these fibers not only reach the tip but may also extend up to 5 mm further anteriorly [5, 19, 24]. The risk of quadrantanopsia due to the location of these fibers along the lateral ependymal wall of the ventricle is one of the primary risks of transcortical approaches to the MTL.