Due to their rarity and ambiguous terminologies, the true identity of PCMs has only been recognized by Horiguchi et al. in 2008 (2). Later, Sodhi et al. clarified the terminology of lesions in the region of dorsum sellae (DS) and separated PCMs (eccentrically placed meningiomas centering on PCPs) from DS or upper clival meningiomas (centrally located meningiomas between PCPs) (4). On the basis of this definition, there have been only 19 reported cases of PCMs including this study. These patients had a mean age of 54 at the time of diagnosis and most of them were female (71.4%) ( Table 1 ).

Anatomically, the PCP is a bony prominence at the superolateral aspect of the DS. Anteromedial to the PCP is the PS. Anterolateral to this process is the posterior edge of the cavernous sinus. The ICA ascends into the posterior aspect of the cavernous sinus where the posterior bend is formed lateral to the PCP. The bifurcation of basilar artery lies posterior to the PCP and DS. The CN III enters the edge of the tentorial dura lateral to the PCP in the oculomotor triangle (12, 13). Thus, meningiomas arising from PCP usually compress the PS anteriorly and the CN III laterally or infero-laterally, and encase the C1-2 segment of the ICA or its branches. The optic chiasm can be even shifted superiorly or antero-superiorly as the tumor expands anteriorly. Therefore, headache, visual disturbance, and diplopia are common presenting symptoms. Huge PCMs also cause facial pain or akinesia if the CN V or ICA perforators is involved (14).

The diagnosis of PCM can be challenging. Radiologically, the “dural tail” sign is seldom detected since the tumor attachment at PCP is very small. Owing to clinical and radiological similarities, some PCMs are preoperatively misinterpreted as DS or upper clival meningiomas, and may be even confused with anterior clinoid meningiomas (ACM) and parasellar meningiomas. The direction of ICA displacement may help in differentiating PCM from ACM, since the ICA is often shifted posteriorly in the case of ACM but anteriorly in the case of PCM (15).

Like most PCMs, the tumor in this case grew upwardly and forwardly beyond the tentorial incisura and up-elevated the right optic chiasma, which could account for progressively visual loss in right eye. However, this case is unique in the tumor site presented unlike those reported in previous studies. Intraoperatively, the tumor was located between CN III and CN IV in the mediolateral direction. Correspondingly, the origin of PCM was found to occupy the junction of the PCP and petrous apex. Thus, the protection of CN IV should also be emphasized during surgery.

The principles of PCM surgery are early devascularization of tumor and careful dissection of tumor from ICA perforators, cranial nerves, PS, and even hypothalamus (1). Several approaches have been implemented to treat this lesion with acceptable risks, and many of them represent variations of the classical pterional, subtemporal, or presigmoid routes with modifications in bone resection. According to surgical directions accessing the PCP, they could be classified into anterolateral, lateral, posterolateral, and complex routes ( Figure 4 ).

The transcavernous-transsellar approach was proposed for PCM resection by Dr. Dolenc in 2003. During surgery, the sphenoid wing, ACP, and the lateral wall of the optic canal are excised extradurally. Then, with additional resection of the PCP and DS, the PCM could be devascularized at the attachment and then excised (15). The classical pterional approach allows early identification of ICA perforators and renders direct access to the tumor through the optico-carotid and carotico-oculomotor corridors. Common disadvantages for these anterolateral routes include hindrance of direct approach to tumor by anteriorly shifted ICA and risks to ICA perforators during devascularization and debulking of the tumor through the narrow spaces between them (5, 15). In addition, early detection of CN III is difficult in the transcavernous-transsellar approach (15). Hence, only soft or small PCMs might be treated with these approaches. So far, only one case of reported PCM has been resected via fronto-temporal craniotomy with wide splitting of the Sylvian fissure (5).

The inherent drawbacks of anterolateral approaches could be circumvented by lateral or posterolateral approaches. In 2013, Shukla et al. firstly reported two cases of PCM excision via transzygomatic subtemporal approach (3). This lateral approach is technically the simplest approach for exposure of tentorial notch and upper clivus. A further incision of the tentorium may be required for a wider exposure. It is a better choice for early devascularization and debulking, and allows for direct access to the PCP between the CN III and posterior communicating artery after resection of small compartment around these structures. Major disadvantages of this approach are temporal retraction and injury to the vein of Labbe, which might be avoided by placing a lumbar drain preoperatively (3).

In 2009, Goto et al. described their experience with five cases of PCM resected via presigmoid transpetrosal approach. This posterolateral route guarantees earliest coagulation of feeding arteries of the PCM. Tracing the CN III from the normal proximal side in the interpeduncular cistern to the involved distal side could prevent unexpected injury of the nerve. This approach also enables safe dissection of the antero-superiorly displaced ICA under direct vision and provides direct exposure to the inferior surface of the optic chiasm, PS, and hypothalamus from the posterolateral side. However, burdensome bone drilling with related risks to hearing and facial nerve function as well as cerebrospinal fluid leakage makes it somewhat less desirable surgical choice. Operational complexity also restricts the popularization of this technique (1). In 2017, Nanda et al. reported a case of resection of a large PCM with obvious brainstem compression via retrosigmoid approach. Although the most amazing finding was the extent of visualization around the PCP region gained through this approach, the narrow width and great depth of this working corridor pose significant risks of neurovascular injury (7).

Size, consistency, and tumor micromorphology are the most important determinants of the extent of resection and outcome of PCM. Small, soft, and well-delineated tumors can be excised totally with any of the above approaches without significant neurological sequelae, while tumors with large size, hard texture, rich vascularity, and multiple micro-lobules encasing perforators often result in incomplete resection (3). Among the 19 reported PCM cases, total resection was achieved in 9 patients (47.4%) ( Table 1 ). It is worth noting that removal of far anterior extension of PCM could not be realized through lateral and posterolateral routes. Under this condition, a basal frontotemporo-orbito-zygomatic orbitozygomatic (FTOZ) approach with wide Sylvian fissure splitting has been recommended. Compared with the classical frontotemporal craniotomy, this approach provides additional anterior temporal or subtemporal route. This surgical trajectory allows direct access to the PCP posterior to the perforating vessels for tumor devascularization and debulking without temporal retraction. Thereafter, the devascularized tumor could be removed through the transsylvian route (4, 5). For extremely huge and complex PCMs, a combined and multi-staged surgical approach could be options of tailor-made surgical strategy. In one reported case, a presigmoid transpetrosal approach was recommended for the first stage combined with an orbitozygomatic transsylvian approach for devascularized residual tumor in the second stage (1).

The ELSI approach was firstly described by Spetzler et al. in 2000 as a distinct variant of the median SCITA approach for accessing the posterolateral mesencephalon (16). Later, it proved its versatility and clinical practicality in treating tumors residing in centrally-located intra-axial structures like the splenium, pulvinar, brainstem, and mesial temporal lobe as well as skull base extra-axial tumors like petroclival meningiomas (9). With the rapid development of neuroendoscopy, the infratentorial space has been recognized as another optimal endoscopic operating area. Earlier anatomic studies have addressed the possibility of endoscope-assisted or purely endoscopic ELSI for accessing the posterior and posterolateral incisural space (17–19). Clinically, endoscope-assisted or purely endoscopic ELSI has been used for removal of supratentorial lesions, such as cavernous malformation in the posteromedial temporal lobe and petroclival meningiomas extending into the middle fossa, with the aid of tentorium incision (10, 20). Based on these studies, we attempted to treat central skull base lesions through this novel approach and have succeed in resection of retroinfundibular CP in a recently reported case (11). Herein, we represented a case of PCM resected via the EF-SCITA approach. Despite with the longest working distance, endoscopy extends our eye to see the central skull base and our hands to excise lesions centering over the PCP.

This posterolateral approach necessitates tentorial disconnection to traverse the tentorial incisura and arrive at the retrosellar space, and requires working between neural structures for PCM removal in the ambient cistern. The decompression of infratentorial tumor facilitates the exposure and resection of the part in the supratentorial space. As with transpetrosal approach, this approach allows tracing the natural trajectory of CN III and IV for nerve protection. Based on our prior experience in resection of retroinfundibular CP, this approach also provides direct exposure to the branches of ICA (including the PCA and anterior choroidal artery), optic chiasm and PS from the posterolateral side (11). Thus, this novel approach might be a promising alternative approach for resection of suprasellar or retrosellar lesions.

Compared with the retrosigmoid view, the supracerebellar view supplies the corridor from the medial to cranial nerves and significantly reduces the manipulation of neurovascular complex (21). By exploiting the natural infratentorial space, this corridor is devoid of cortex retraction seen in frontotemporal craniotomy. This approach also avoids other inherent shortcomings of above-mentioned anterolateral, lateral, and posterolateral routes, such as hearing loss, damage to the central skull-base bone, and injury of the Labbe vein. In addition, this technique simplifies craniotomy procedures, with a small incision that does not affects the appearance of patients (11). Another important advantage of this approach is high-definition wide-angle visualization and close-up observation endowed with endoscopy, which further minimize injury especially to various neurovascular structures behind the DS. Concerning the small size of the tumor, only a forward viewing endoscope was utilized in this case. The adoption of angled endoscope will allow visualization of anatomical structures out of the axis of the telescope and minimize the chance of tumor remnants under the condition of large PCMs.

Despite these merits, we are clearly aware that it is not without flaws. Firstly, endoscopy technique has difficulties when handling deep bleeding and necessitates a steep learning curve. Secondly, it is not entirely risk free because it may endanger the brainstem and cerebellum, and has inherent shortcomings of suboccipital craniotomy such as injury to the cerebrovenous system (including the petrosal venous, transverse sinus, and sigmoid sinus). Thirdly, unlike lateral and other posterolateral routes, tumor devascularization could only be achieved after tumor removal owing to the final exposure of PCP. Thus, bleeding control should be stressed especially in resection of massive or easily bleeding tumors. Last but not least, this case represented resection of a small PCM. Considering the surgical corridor, it would be an ideal approach for PCMs with lateral extension to the temporal lobe or posterolateral extension to the petroclival region, but may not be enough for complex PCMs with extreme suprasellar extensions. In this situation, microscopic FTOZ craniotomy may be needed either as an add-on or instead of the EF-SCITA. We believe multi-corridor hybrid surgery via combined EF-SCITA and microsurgical middle fossa approach might be an ultimate piece of the surgeon’s armamentarium to improve outcomes in patient with complex DS lesions.