Detailed Arterial Anatomy and Its Anastomoses of the Sphenoid Ridge and Olfactory Groove Meningiomas with Special Reference to the Recurrent Branches from the Ophthalmic Artery

Anatomy of the Dural Arteries

Martins et al¹ reported the detailed microsurgical anatomy of dural arteries as follows: The supply to the middle fossa and paracavernous dura derives laterally from the MMA, accessory meningeal artery, and ascending pharyngeal artery and medially from the intracavernous branches (ILT) of the ICA. They also described the contributions to these regions from the recurrent branches of the ophthalmic and lacrimal arteries. There is confusion about the definition of the recurrent ophthalmic arteries. Lasjaunias et al⁶ reported that both the superficial and deep recurrent ophthalmic arteries pass backward through the lateral and medial parts of the SOF, respectively. Meanwhile, Martins et al¹ reported that the superficial recurrent ophthalmic artery, deep recurrent ophthalmic artery, and RMA pass through the optic canal, the medial part of the SOF, and the lateral part of the SOF, respectively. Therefore, descriptions of both authors were competing. In the present article, we describe the RMA as posteriorly recurrent branches from the OphA passing through the OC or SOF, including the superficial and deep recurrent ophthalmic arteries. Martins et al¹ and Hayreh³ described the recurrent branch passing through the OC. However, to the best of our knowledge, no articles have shown this recurrent artery passing through the OC in imaging or cadaveric studies.

Yamaki et al⁷ reported details of feeding arteries in parasellar meningiomas, including 9 sphenoid ridge meningiomas. In this article, they analyzed 2 lesions (22%) in the sphenoid ridge that were supplied from the superficial recurrent ophthalmic artery or recurrent meningeal artery, respectively. The current imaging technology and slab MIP images provide us with the ability to analyze detailed arterial anatomy. We can demonstrate that RMAs from the OphA were involved in most of the lesions (90%) in the present study.

Anastomosis of Each Feeder

Numerous anastomoses between each artery have been reported.^2⇓⇓-5 In the setting of hypervascular lesions related to the dura, such as meningiomas or dural arteriovenous fistulas, these feeding arteries can be enlarged and tend to connect.⁸ In the present study, there were various anastomoses in 75% of cases, and 4 cases had multiple anastomoses.

According to the embryologic concepts of the OphA, the connection between the OphA and the ILT is the remnant of the dorsal ophthalmic artery and the connection between the RMA and the MMA is the remnant of the supraorbital artery (branch of the stapedial artery).^9⇓⇓-12 De La Torre and Netsky¹³ reported the persistent primitive maxillary artery in the human fetus. They reported that this artery arises from the cavernous portion of the ICA and anastomoses with the lacrimal artery and the anterior branch of the MMA. We can comprehend the complicated anastomoses among the OphA, ICA, and MMA using these findings. Therefore, it is certain that there are various anastomoses among branches of the OphA, ICA, and ECA.

Clinical Implications

Preoperative embolization for meningioma is performed as a standard treatment. Although its usefulness is widely accepted, the embolization of target vessels other than the ECA has been reported as a risk for procedural complications in a nationwide surveillance in Japan.¹⁴ Rosen et al¹⁵ reported their treatment results for preoperative embolization using microparticles to 167 cranial base meningiomas, and 9% of their patients experienced permanent neurologic deficits. They also showed that embolization of the MHT, MMA, and ascending pharyngeal artery involves the risk of a minor or permanent neurologic injury. Meanwhile, Waldron et al¹⁶ reported their treatment results for preoperative embolization using microparticles and coils to 119 cranial base meningiomas, with no complications. Some authors have reported the usefulness of embolization of the OphA branches.^17,18 In limited cases, embolization may be performed for branches of the OphA, but it is associated with a risk of visual disturbance due to migration of the embolic material to the central retinal artery.¹⁷ Appropriate knowledge of anatomy, distal catheterization, and the gentle injection of embolic material without reflux are necessary for safe embolization of the OphA.¹⁹

In our opinion, the RMA-OC is inappropriate as the target vessel of preoperative embolization due to the difficulty in catheterization because of its small caliber and short distance from the origin of the OphA with an acute angle. The RMA-LSOF may have a lower risk because of its origins and anastomoses. The RMA-LSOF usually originates from a lacrimal artery and is anastomosed with the sphenoid branch of the MMA. However, if the RMA-LSOF has an anastomosis with the ILT, it involves a risk of the migration of particulate or liquid materials to the ICA. Thus, we recommend embolization of both of the RMA and the ILT using detachable coils in such a case.

Another precaution related to embolization is residual feeding from anastomotic vessels. If there is an anastomosis between each feeder, proximal occlusion of the feeder, which can be easily catheterized, can result in increased blood flow to the tumor from the residual feeder. One should embolize the common trunk of both feeders or the proximal parts of both feeders (Fig 2).

A detailed understanding of arterial microanatomy is also useful in direct surgery. A large sphenoid ridge meningioma requires skull base techniques, such as the transzygomatic approach.^20,21 Using a skull base technique and extradural drilling of the sphenoid wing to the SOF, we can devascularize the feeder from the ECA and RMA-LSOF/MSOF.²¹ After these procedures, we open the dura mater, and we can debulk the intradural tumor, most of which is already devascularized, whereas, we may not be able to find the feeder from the RMA-OC until identification of the optic nerve. Moreover, coagulation to the RMA-OC involves the risk of optic nerve injury. If we identify all of the feeders using preoperative angiography, we can search for feeders on the basis of the anatomic landmarks of the bone and devascularize them in a safe and efficient manner.

Limitations

The current study has some limitations. First, the number of patients was small. Second, we did not confirm the presented feeding arteries with operative findings. In future studies, we should examine the accuracy of the presented results using operative findings or cadaveric research.