Discussion

Prior to cortical mastoidectomy, it is important to determine whether or not the bone is well-pneumatised. Sclerotic cases involve a narrow mastoid space surrounded by dura, the sigmoid sinus and the posterior wall of the external auditory canal. It is important to avoid damaging these boundary landmarks. Well-aerated cases with thin air cell septa may have thin facial bony canals. Careful drilling is essential to avoid canal damage. During mastoidectomy, it is important to locate the aditus ad antrum, which ensures the absence of iatrogenic trauma. However, a sclerotic mastoid, as well as granulation tissue in the middle ear and mastoid, may hinder identification of the aditus ad antrum.^{Reference Long, Athar, Mahmud and Saim1}

Iatrogenic trauma of the facial nerve principally affects the second genu of the fallopian canal. This may occur when drilling the mastoid toward the aditus. The surgeon may inappropriately drill more inferiorly than the aditus. As the second genu of the facial nerve lies inferior and medial to the aditus, this is the most likely site of injury.^{Reference Long, Athar, Mahmud and Saim1} The semicircular canals may also be accidentally injured when drilling toward the aditus. Injury to the labyrinth can cause sensorineural hearing loss and serious suppurative intracranial infection.^{Reference Long, Athar, Mahmud and Saim1}

Anterior displacement of the sigmoid sinus may lead to damage during mastoid drilling. An anteriorly located sigmoid sinus often narrows the mastoidectomy boundary, preventing the surgeon from reaching the antrum because of the limited field of view.^{Reference Çam and Karataş2} When creating the second gate via posterior tympanotomy, it is essential to identify the facial canal (to avoid damaging the facial sheath) and the chorda tympani. The latter structure can be located via soft drilling with a large diamond burr under irrigation.

Paralysis of the facial or chorda tympani nerve is a potential complication of the facial recess approach; such complications are rare but serious, and their likelihood can be minimised by obtaining a full understanding of facial recess anatomy. The facial recess length and width, and the angle between the facial and chorda tympani nerves, are larger in pneumatised temporal bones than in sclerotic and diploic bones. Facial recess length varies greatly by the type of pneumatisation. Variations in the origin of the chorda tympani branch and pneumatisation pattern affect the size of the facial recess.^{Reference Jain, Deshmukh, Lakhotia, Kalambe, Chandravanshi and Khatri3} The round window niche is not always fully exposed by posterior tympanotomy. Öztürk et al. found that complete round window exposure through the facial recess was possible for most temporal bones (79.2 per cent).^{Reference Öztürk, Göde, Çelik, Orhan, Bilge and Bilgen4}

For the third gate (created via cochleostomy or the round window approach), pre-operative CT evaluation of round window shape and position, cochlear ossification, and any small dimensions or narrow turns, is very useful.^{Reference Karkas, Champfluer, Uziel, Mondain, Puel and Venail5} Thinning of the posterior external auditory canal wall may improve round window visibility. The round window varies markedly in size and shape.

The role of the round window as a portal for electrode implantation is well supported by the fact that it is the safest landmark for identifying the scala tympani. Hamamoto et al. used a pin gauge to classify the topographical relationship between the facial and chorda tympani nerves and posterior wall of the external acoustic canal into five types.^{Reference Hamamoto, Murakami and Kataura6} It can be difficult to find the round window if the middle-ear mucosa is thickened by chronic otitis media, anatomically changed by prior surgery or obstructed with a pseudo-membrane; gentle dissection using micro-instruments may be required.

Factors that influence the optimal trajectory for atraumatic electrode insertion include the orientation and size of the cochlear basal turn, and the anatomy of the hook region that lies immediately behind the round window membrane.^{Reference Luers, Hüttenbrink and Beutner7} Several aspects of round window anatomy are relevant to its use as a portal for electrode insertion. Certain anatomical features in fact suggest that the round window is not the ideal portal for cochlear implantation. An electrode inserted straight through the round window will be aimed directly at the modiolus.^{Reference Roland, Wright and Isaacson8} If the posterior-superior overhang of the niche is not drilled away, it is almost impossible to insert an electrode into the scala tympani alone; instead, it will be directed toward the modiolar wall and spiral ganglion.^{Reference Roland, Wright and Isaacson8} In one study, cochlear electrodes were more likely to cause osseous spiral lamina trauma when placed in the scala vestibuli via round window implantation rather than via cochleostomy.^{Reference Zhou, Friedmann, Treaba, Peng and Roland9}

The goal of cochleostomy is to enter the scala tympani as near as possible to the centreline or main axis, to avoid the need for drilling of, or application of direct force to, the spiral ligament region of the outer and superior walls.^{Reference Adunka, Radeloff, Gstoettner, Pillsbury and Buchman10} Thus, an ideal insertion vector minimising the force delivered to, and subsequent damage of, intracochlear structures is required.^{Reference Meshik, Holden, Chole and Timothy11,Reference Breinbauer and Praetorius12} Electrodes can become displaced even in patients with normal inner-ear anatomy. In particular, an electrode may be inappropriately directed into the superior or lateral semicircular canal through the vestibule^{Reference Jain, Deshmukh, Lakhotia, Kalambe, Chandravanshi and Khatri3,Reference Breinbauer and Praetorius12} or internal auditory canal, or even the internal carotid artery. Electrode misplacement is usually caused by drilling along an inappropriate axis. A microscope should be used to identify the round window niche and membrane. Computed tomography of a cochlear duct lying under the promontory of the middle-ear cavity is very helpful.

The internal carotid artery and jugular bulb are at risk during cochlear implantation. Therefore, the access route to the cochlear lumen should be chosen before surgery.^{Reference Sahni, Singla, Gupta, Gupta and Aggarwal13} A high-riding jugular bulb may cover the round window or basal cochlear turn, thereby hindering access to the cochlear lumen via the standard transmastoid approach.^{Reference Di Lella, Falcioni, Piccinini, Iaccarino, Bacciu and Pasanisi14} Hara et al. found that reconstructed 3D-CT images were useful when surgically planning cochlear implantation in a patient exhibiting an abnormal facial nerve course;^{Reference Hara, Takahashi and Kanda15} 3D images are optimal for assessing temporal bone anatomy, and they show the skull base and any middle- or inner-ear anomalies.^{Reference Joshi, Navleka, Kishore, Reddy and Kumar16,Reference Nakamura, Dias, Sasaki, Sugimoto, Hatano and Ito17} Breinbauer and Praetorius^{Reference Breinbauer and Praetorius12} described the optimal spatial orientation of insertion vectors for cochleostomy and the round window approach through 3D reconstruction of the temporal bones of 50 cochlear implant candidates (100 ears).

High-resolution CT is the only imaging modality that can pre-operatively detect abnormalities, thus minimising the risks associated with an abnormally located facial nerve or narrow recess, for example.^{Reference Karkas, Champfluer, Uziel, Mondain, Puel and Venail5} Two-dimensional imaging lacks the power of 3D-CT. Round window visibility can be assessed by rotating the 3D reconstructed image after virtual mastoidectomy and posterior tympanotomy. Three-dimensional CT also reveals important features of the temporal bone, thus allowing surgeons to identify and address any difficulties expected during implantation.