The use of subtotal petrosectomy in cochlear implant candidates with chronic otitis media

The cochlear implant surgeon facing the decision to implant a patient with chronic otitis media (with or without cholesteatoma) or a patient with failed previous tympanomastoid surgeries has to consider the following realities:

There have been several techniques proposed in the literature to manage patients with chronic otitis media requiring a cochlear implant. They can be divided into three groups:

(A) “Covering techniques” to avoid electrode extrusions by wrapping the electrode cable into dense patient’s own tissues: Schlondorf et al. [11] used full thickness postauricular skin and soft tissue flap to cover the electrode in the mastoid cavity. Manrique et al. [12] suggests using tragal cartilage and fascia to cover and hold the electrode in the cavity. Others suggest reconstructing of the tympanic membrane and posterior ear canal wall [13]. Kojima et al. [14] used a canal wall reconstruction technique with mastoid obliteration in 2 patients with open cavities. However, Olgun et al. [15] reported 37 cases with COM and existing open cavities or middle ears converted to open cavities prior or at the time of cochlear implantations. In 7 (19 %) of them the electrode cable disrupted the epithelial lining of the cavity and reimplantation was necessary. El-Kashlan et al. [16] recommended the closure of external ear canal and leaving intact the pneumatised mastoid that would enable later serial CT follow-up. Roehm and Gantz [17] presented a case of chronic otitis media that developed after cochlear implantation and resulted in the need of explantation. They also reviewed the literature from 1995 to 2004 for papers describing complications of cochlear implantations in COM patients. In 14 from 100 patients a reoperation became necessary and in 7 cases the implant had to be removed due to complications. None of those seven patients was operated with our SP technique and in fact two of them required SP as a treatment of previous complications.

All those techniques carry the lifetime risk of electrode exposure and extrusion with subsequent need to remove the implant. They also do not guarantee a solution against recurrent or residual disease and in case of inner ear malformation a CSF leak would pose a serious problem. We therefore strongly discourage from using these “covering” techniques.

(B) “Bypass techniques” using a surgical approach for electrode insertions away from the diseased middle ear and mastoid: Kojima et al. [14] used a transcanal approach drilling the grove in the posterior canal wall in patients with adhesive otitis to avoid drilling in a previously infected and scarred mastoid. Olgun et al. [15] advocated the use of a subfacial approach in those cases with previous modified radical cavities and reported stable results in 13 patients with a follow-up from 1 to 5 years. Colletti et al. [18] reported a method of electrode insertion via the middle fossa approach in cases with chronic otitis media. This technique requires a craniotomy, carries higher risk of facial nerve injury and leaves the middle ear disease unsolved. In patients with open cavities the risk of continuing recurrent discharge from the ear remains. Again, we strongly discourage from using any of these “bypass techniques”, since the underlying problem is not solved.

(C) “Subtotal petrosectomy” technique to eliminate the chronic middle ear and mastoid disease and implantation of electrodes into a clean surgical field: Fisch and Mattox [5] described the detailed technique of SP already in 1988 and 10 years later he introduced the concept for patients in need of a cochlear implant. Among the first five patients were patients with temporal bone and inner ear malformations, meningocele and chronic otitis media [7]. It took another 10 years until cochlear implant surgeons became familiar with this concept and it is only now, that we can refer to various centres reporting their experience in the ENT literature. The Antwerp group summarized 29 patients with severe chronic otitis media resistant to medical or previous surgical treatments who underwent an SP. None of the patients had recurrent otorrhea, one had a residual cholesteatoma and one patient out of five who received a CI in a single-stage procedure revealed a severe complication: he had two consecutive flap failures with wound breakdown 6 and 5 months after revision surgeries and finally required explantation and re-implantation [19].

In the group of 32 patients operated by means of SP combined with cochlear implantation at Piacenza reported by Free et al. [6], 4 had chronic suppurative otitis media and 13 patients had previous canal wall down surgeries (one with an electrode extrusion through the retroauricular skin of the modified open cavity). One patient (3 %) in the SP group developed a retroauricular wound infection with granulation tissue requiring reoperation and repositioning of electrode array. Our group of 19 patients had a follow-up until 10 years with no early or late complications. Postelmanns et al. [20] described a group of 13 patients with COM out of 156 (8.3 %) patients that received cochlear implants in Maastricht. In eight patients in this group SP was performed without any complications. In the remaining five patients one major complication occurred requiring explantation of the CI. The patient had simple myringoplasty for a dry perforation and developed skin infection leading to reoperation and finally to the explantation of the device. They advocate staged procedure—first SP and after 3–6 months cochlear implantation—in any case of active disease, including any case with cholesteatoma.

More recently, Bernardeschi et al. [21] reported 24 patients with chronic otitis media treated with single-stage cochlear implantation and subtotal petrosectomy from the group of 30 cases treated with this technique. They strongly advocate one-stage surgery in all patients with prolonged antibiotic therapy in case of positive bacterial culture. They reported no complications except two abdominal hematomas nor cholesteatoma recurrence; however, the follow-up in six of the reported cases was only 3 months.

We strongly favour the SP technique for patients with chronic otitis media, previous modified radical cavities or failed tympanomastoid surgeries who require a cochlear implant procedure. The SP allows maximum exposure of the temporal bone, reveals a high chance of radical removal of any disease (minimal risk of residual disease) and avoids any recurrence of the disease (e.g. further middle ear atelectasis is not possible since the drum was removed), and limits the risk of meningitis in cases of additional inner ear malformations. Although we have not specifically addressed the hearing outcome, generally speaking, the functional outcome was similar to cochlear implants in healthy middle ears, as observed also in the Antwerp and Piacenza group of patients [6, 20].

The decision in these cases is therefore not, if there are other alternative techniques to the SP, but whether to stage the procedures or to perform a single-stage implantation.

In 14 (74 %) patients from our group a single-stage surgery was performed. Obvious advantage of this strategy is that patient has just one surgery and in 3–4 weeks the cochlear implant is activated. Because of the risk of biofilm formation on the implant, in patients with active discharge from the ear with multiresistant bacterias, extensive cholesteatoma or in previously irradiated temporal bones, staged surgery may be necessary. Disadvantage of single-stage surgery is that imaging with diffusion—weighted MRI is not easily possible anymore after cochlear implant placement (or the magnet needs to be removed), making an early diagnosis of residual (not recurrent) cholesteatoma impossible. In three of our patients cholesteatoma was present and in two of them single-stage surgery was performed. In one case the procedure was staged and second stage performed after non-EP DW MR excluded residual disease. We suggest performing imaging in all two-stage cases 6 months or longer after the first stage. The period between stages should be at least 6 months as non-EP DW MR imaging technique is able to detect pearls of cholesteatoma larger than 2–3 mm depending on the study protocol and class of the MR equipment [19, 22]. In case of a positive finding, this small residual perl can be extracted at the same time as the staged CI implantation takes place. As we are aware that small pearls or a flat residual cholesteatoma matrix may not be detectable at 6 months postoperatively, we suggest that an MRI is optional and not mandatory. If the period is too short small residual pearls enlarging in the obliterated cavity may be missed radiologically and even during second stage surgery, since the fat pad is not completely removed at the second stage. Therefore, careful surgical inspection of the cavity should be performed during the second stage. In patients operated with a two-stage technique the fat obliterating the cavity is partially removed allowing the identification of the previous bony margins of the cavity, the promontory, round and oval window niches. After cochlear implant insertion the cavity is further filled with freshly harvested subcutaneous fat. In all our patients perioperative antibiotic coverage was used in both stages to further reduce the risk of implant infections.

Recently Vincenti et al. [23] reported long-term results of 12 patients after subtotal petrosectomy and single- or two-stage cochlear implantation in patients after open cavity surgery. They reported one patient with residual cholesteatoma and one with wound breakdown at the external meatus. They also advocate follow-up HRCT about 1 year after surgery and further imaging depending on clinical symptoms. The use of subtotal petrosectomy and implantation of middle ear implants has also been reported [24].

Looking at our own series and experience and reviewing the literature, we suggest the following treatment algorithm (Fig. 3).

Main indications are patients with chronic otitis media with or without previous surgeries, CI candidates with inner ear or temporal bone malformations limiting the surgical exposure (posterior tympanotomy) or harbouring the risk of CSF leak and meningitis including also transverse temporal bone fractures. The surgeon has to decide upon a primary or staged implantation. A follow-up CT scan (or Cone beam CT) performed 1 year after cochlear implant insertion confirms the air tight closure of the Eustachian tube (no air in the protympanum), identifies the position of the electrodes within the cochlea and in case of doubt of residual disease can be used as a baseline scan for comparison 3–5 years later. Pathologies within the fat pad may be identified due to differences in tissue density in Hounsfield unit scale. Furthermore, serial CT examination may show growth of the lesion within the obliterated cavity suggesting residual cholesteatoma [6].

The only contraindication would involve a CI-candidate with an attempt for electroacoustic amplification. However—as mentioned above—this is very unlikely due to the disease process both in the middle ear and its previous effects on the inner ear.