Keywords
cochlear implant - mastoidectomy - facial nerve
Introduction
The posterior tympanotomy approach for cochlear implant (CI) surgery, described by
House, has been the most commonly used CI surgery worldwide. So far, reports of facial
nerve injury during facial recess surgery for CI show a consistent rate of less than
1%, with several studies reporting a rate of 0.7%.[1]
[2]
[3]
[4]
[5]
[6]
[7] The vast majority of these are partial weaknesses of brief duration or delayed pareses
of a temporary nature. Nonetheless, uneasiness of some surgeons drilling near the
facial nerve has led to the development of alternate techniques for CI[1].
Some authors consider these alternative techniques (non-mastoidectomy approaches)
critical and helpful in the presence of anatomic constraints, such as a small mastoid
cavity, where there is greater difficulty in a facial recess approach. Additionally,
there is a significant chance of aberrant facial nerve location in cases of labyrinthine
dysplasia. Hoffman et al.[2] reported facial nerve anomaly in 16% of patients with CI. Moreover, one third of
these patients had either common cavity malformation or hypoplastic cochlea with aberrant
facial nerve course.[2]
Non-mastoidectomy CI approaches include the suprameatal approach (SMA) and its modifications,
the transcanal approach and its modifications and the pericanal approach for electrode
insertion.
Methods
We searched several medical databases, including LILACS, MEDLINE, SciELO, PubMed and
the Cochrane Library in February 2015 to find out relevant articles.
We focused our review on studies involving the non-mastoidectomy approaches to CI.
Following Júnior et al.,[8] surgical complications were classified into major (if they require additional surgery
or hospitalization), and minor, (when they resolve with outpatient treatment or even
with no treatment). Major complications involve meningitis, flap necrosis, device
failure, electrode extrusion, facial nerve paralysis and others; while the minor complications
involve facial nerve stimulation, electrode migration, vertigo, tinnitus, and others.
The major complications are less common.[8]
The authors analyzed the results of 12 studies that use the various non-mastoidectomy
approaches to CI, highlighting their importance as alternatives to the traditional
mastoidectomy.
The Pericanal Approach for Electrode Insertion
The Pericanal Approach for Electrode Insertion
The pericanal approach is one of the alternative surgical techniques for CI, which
allows electrode insertion without performing a mastoidectomy. In this procedure,
the electrode is placed directly through the EAC into the tympanic cavity. In brief,
the procedure begins with a standard retroauricular incision, whereby the skin in
the posterior EAC is lifted until the annulus of the TM is reached. The TM is retracted
anteriorly, exposing the middle ear and its contents. A standard cochleostomy is performed
slightly anterior to the round window. Then, using a 1.6 mm diamond burr, a vertical
groove of ∼2 mm is created from the posterosuperior region of the bony EAC, above
the incus body toward the lateral EAC, where a short superficial tunnel is created.
A standard seat for the implant is created and fixed in the parietal bone; then, the
electrode is led through the tunnel and the vertical bony groove into the tympanic
cavity, where it is inserted into the cochleostomy. The electrode is fixed in place
with glass ionomer cement and the groove is filled. The skin flap is replaced along
the EAC and the ear is packed with gauze strips for 7–10 days.[7]
The Suprameatal Approach (SMA) and Its Modifications
Kronenberg et al. originally developed the Suprameatal approach to CI surgery in 1999,
based on a blind atticotomy technique which was similar to a widely accepted approach
to cholesteatoma.[3]
[4] The procedure involves raising a retroauricular skin flap and creating subperiosteal
flap to drill a seat in the temporoparietal bone for the receiver-stimulator. Through
the postauricular incision, a tympanomeatal flap is elevated. Once the chorda tympani
nerve is identified, a 1-mm groove is drilled in the scutum posterosuperior to the
facial nerve and lateral to the incus until the incus short process is exposed. A
blind tunnel is then drilled in the parietal bone just inferior to the temporal fossa
bony plate and superior to the external auditory canal (EAC), aiming for the groove
created in the scutum. Then, the cochleostomy is performed through the EAC, and the
electrode is conducted through the suprameatal tunnel and the groove into the middle
ear. It is then redirected to the cochleostomy in a near-vertical direction, and,
again, redirected in an antero inferior direction in line with the scala. Later, Postelmans
et al. introduced modifications to the original technique, in which the middle ear
cavity is entered via an endaural tympanotomy rather than a postauricular incision,
creating a subperiosteal tunnel between the seat for the receiver-stimulator and the
suprameatal tunnel to protect the electrode.[5]
In 2004, Kronenberg and Migirov described 35 patients aged 1 to 67 years with ∼8 months
follow-up period that underwent CI by SMA[9]. Electrodes were inserted through Antro-inferior cochleostomy using several devices.
They reported no surgical complications[9].
Two years later, Kronenberg et al.[10] described 188 patients who underwent SMA for CI. They reported no facial nerve injuries
or cerebrospinal fluid (CSF) leaks and no cases of electrode extrusion into the EAC.
There were, however, 4 perforations of the tympanic membrane (TM) (2%) where the scutum
was drilled.[10]
In 2004, there were 255 reported cases of patients who underwent CI by SMA, where
185 were in Tel Aviv and 70 in Vienna. Electrode insertion was through Antro-inferior
cochleostomy with Combi 40 and Combi 40+ (MEDEL) devices and Nucleus and Clarion electrode
devices. There were no surgical complications reported[11].
In China, 2008, Yin et al.[12] described 45 patients who underwent CI by SMA (a total 47 ears). Patients average
age was 5.7 years; 29 were male and 16 female, undergoing a follow-up period ranging
from 1 to 20 months. Electrodes were inserted through Antro-inferior cochleostomy.
Of the implants, 42 were Med-El C40+ while the other five were Clarion 90k. They reported
that 38 patients with Med-El implants had all electrode pairs entirely inserted in
the cochlea; one patient with cochlear ossification, however, had 9 pairs of electrodes
inserted, and another patient with profound cochlear dysplasia had 8 pairs of electrodes
inserted. All electrode pairs were inserted in the five patients using a Hires 90KTM
implant. During surgery, a ‘‘gusher’' occurred in one patient with profound cochlear
dysplasia, and was treated by sealing the cochleostomy using a small piece of teamporalis
muscle and biological glue. The electrode was inserted lateral to the chorda tympani
in one patient because the chorda tympani had adhered to the body of the incus.[12] Three patients with large vestibular aqueduct syndrome (LVAS) and one patient with
profound cochlear dysplasia had postoperative vestibular dysfunction symptoms such
as vertigo, nausea, and vomiting. Among the 30 patients who were followed for more
than 6 months, 12 had open set speech perception and could effectively communicate
with others; 14 patients learned to speak short sentences even if not clearly; and
4 patients could only say single words.[12]
In 2012, Tange and Tange[13], from the Netherlands, reported 260 cases with mean age 39.6 years who underwent
CI by SMA. The electrodes were inserted through Antro-inferior cochleostomy. The study
used four different types of cochlear implants: the Nuc1eus 24-Countour, the Nuc1eus
24-Countour Advance with Softip, Advanced Bionics HiFocus Helix and Medel Sonata.
The authors found that the minor complication rate was 23%, which consisted of tinnitus
(7.2%), postoperative vertigo (5.2%), TM perforation (1.3%), hematoma (1.3%), and
other causes (3.8%). In one patient (otosclerosis), mild facial nerve stimulation
was found and could be managed by switching off several electrodes. All of the five
patients that reported minor complications were successfully treated. On the other
hand, all five cases of major complications (3%) developed postoperatively. The major
complications consisted of two cases of implant extrusion due to wound infection,
a fausse route of electrode 11, a misdirected electrode in severe otosclerosis, and
a case of explanation due to psychiatric illness and pain sensations. All these five
cases were re-implanted and had successful results thereafter, except for the patient
with the psychiatric illness. This patient continues to complain, in spite of a perfectly
functioning device and fully healed wound. One device failure in one patient (speech
recognition had declined one year after implantation) required explanation and re-implantation.[13]
In 2010, Guevara et al.[14] described 23 patients who underwent CI by SMA in France, of which 13 were female
and 10 were male. Patient ages ranged from 8 months to 63 years with mean follow up
duration of 22 months. Electrodes were inserted through Antro-inferior cochleostomy.
In the first patient, the anterior facial recess hole was done too high; thus, the
incus had to be removed to allow the electrode through. Although the tympanomeatal
flap had been elevated without incising the posterior wall of the external auditory
skin, no TM was observed.[14]
In the same year, Postelmans et al. from the Netherlands described 108 patients who
underwent CI by SMA, with mean age 39.6 years and mean follow-up period of 25.7 months.
Electrode insertion was through Antro-inferior cochleostomy using Nucleus 24 Countour,
Nucleus 24 Countour Advance with Softip, and Advanced Bionics HiFocus Helix devices.
The study reported an overall major complication rate of 3.7% (4 out of 107 patients)
that developed postoperatively. Results included extrusion of the implant due to wound
infection (n = 2), a wrong route for the electrode (n = 1), and device failure (n = 1). The minor complication rate was 23.4% (25 out of 107 patients).[5]
In 2012, Zernotti et al[15] described 80 cases of CI by SMA in Spain and England, in which electrodes were inserted
through Antro-inferior cochleostomy. The authors reported that one patient suffered
from a wound infection, four suffered from vestibular complications and two suffered
from electrode displacement. As for major complications, there was only one case documented,
in the form of device failure.[15]
Therefore, of the 1014 patients with SMA CI whose cases have been published, there
were 99 (9.8%) minor and 13 (1.3%) major complications reported and no reports of
facial nerve paresis or paralysis ([Table 1]).
Table 1
The complication rate differences between transcanal and suprameatal approaches (SMA)
|
SMA
|
Transcanal
|
Chi-square
|
P value
|
|
Total number of CI
|
1014
|
266
|
|
|
|
Total complications
|
112 (11%)
|
24 (9%)
|
0.908
|
0.34 NS
|
|
Major complications
|
13 (1.3%)
|
5 (1.9%
|
0.543
|
0.46 NS
|
|
Minor complications
|
99 (9.8%)
|
19 (7.1%)
|
1.843
|
0.17 NS
|
|
Electrode extrusion
|
8 (0.79%)
|
5 (1.88%)
|
2.494
|
0.114 NS
|
Abbreviations: CI, cochlear implant; NS, non-significant; SMA, suprameatal approach.
The Transcanal (Veria) Approach
The Transcanal (Veria) Approach
Transcanal operation can be summarized by the following steps: 1- endaural or retroauricular
approach to the middle ear with elevation of a standard tympanomeatal flap, 2- inspection
of the middle ear anatomy (cochlea, fallopian canal, round window niche), 3- straightening
of the posterosuperior bony canal wall, 4- cochleostomy, 5- drilling the suprameatal
hollow, 6- drilling the transcanal wall direct tunnel with alignment to the cochleostomy,
7- extension of the skin incision and preparation of the skin flaps and subperiosteal
flap, 8- creating the bed and fixing the device, 9- insertion of the electrode, 10-
manipulation of the excess electrode in the suprameatal hollow, and closing.[6]
In Spain, 2008, Slavutsky and Nicenboim[16] described 10 patients who underwent CI by the transcanal approach, whereby electrodes
were inserted via the round window. The group's mean age was 33.1 years and consisted
of 3 men and 7 women, going through a follow-up period that ranged from 3 to 18 months.
The authors reported that the ten devices had been working properly and results were
as expected for all patients, without complications.[16]
In 2009, from the Kingdom of Saudi Arabia (KSA), Taibah[17] described 131 patients who underwent CI by the transcanal approach, with electrodes
inserted via Antro-inferior cochleostomy. The group was comprised of 115 children,
with ages ranging from 10 months to 14 years, and 16 adults, with ages ranging from
24 to 58 years, with a male-to-female ratio of ∼1:1. Follow-up duration ranged from
2 to 46 months. The author reported five cases of small TM perforations after acute
otitis media; these TM perforations healed spontaneously after medication in four
children. One adult needed myringoplasty. Furthermore, one child with bilateral CI
who had a polyp in the right meatus of the EAC did not respond to topical antibiotics.
The child underwent a re-exploration of the post-auricular area under general anesthesia,
due to fear of possible electrode extrusion. The polyp was removed and the EAC bone
had completely healed with no extrusion of the electrode array. In two children, an
unexpected infection developed around the receiver in the temporal region.[17]
Mostafa et al.,[18] 2014, described 125 cases in Egypt who underwent CI by the modified transcanal approach.
The study group's mean age was 3.4 years, male-to-female ratio was 2.1:1, and follow-up
period ranged from 18 to 50 months. Electrode insertion was via the round window and
the devices used were Cochlear (Nucleus Freedom), MedEl (Sonata), Advanced Bionics
(His res 90 K), and Neurelec (Saphyr). They reported 5 gushers; 2 patients with EVAS
(enlarged vestibular aqueduct syndrome) and 3 patients with IP2-type malformations.
Six patients suffered chorda tympani injury; two of them complained of loss of taste
on the ipsilateral tongue. Two cases had a small TM perforation; both were eventually
grafted uneventfully, without compromising device activation. Two patients showed
electrode exposure, one of which required revision. One patient had severe infection
with extrusion of the device one year after its successful implantation.[18]
In sum, of the reported 266 cases of transcanal CI, 24 suffered complications, of
which 19 (7.1%) were minor and 5 (1.9%) were major, with no reported facial palsy
or paralysis ([Table 1]).
The differences in complication rates between transcanal and SMA is non-significant
([Table 1]).
The Pericanal Approach for Electrode Insertion
The Pericanal Approach for Electrode Insertion
Hausler 2002[19], in Thailand, described 15 patients with an age range of 10 to 48 years, (8 females
and 7 males), with follow-up period ranging from 6 to 24 months, who underwent CI
with pericanal insertion technique. Combi 40 (MEDEL) was used in 14 patients and a
Nucleus CI24M in one patient. The author found that in one patient with a narrow space
between the long incus process and the malleus handle, the incus had to be luxated
and displaced posteriorly to allow easier passage of the electrode through the superior
part of the tympanic cavity. In four patients, small lacerations of the skin of the
EAC occurred during elevation of the posterior tympano-meatal flap. The EAC healed
without problems in all cases. In two patients, a small postero-inferior lesion of
the TM had to be closed at the end of surgery with a mini-graft of adipose tissue.
Despite some operative difficulties and problems, there was no reported post-operative
complication.
Discussion
From the literature analyzed, we found that among the 1014 patients who underwent
CI by SMA, 112 had surgical complications (11%), 13 (1.3%) had major complications,
and 99 (9.8%) had minor complications.
The major complications comprised four cases (0.39%) that had electrode extrusion
due to infection, four cases that had misdirected electrode, four cases of device
failure, and a case that required exploration due to psychiatric illness and pain
sensations.
On the other hand, among the 266 cases that underwent CI by the transcanal approach,
24 (9%) had surgical complications, 5 had major complications (1.9%), and 19 had minor
complications (7.1%).
The major complications were three cases (0.3%) of electrode extrusion due to infection
and two cases (0.2%) with electrode exposure.
Hashemi et al.,[19] in a study of 181 cases that underwent the PTA technique, reported major surgical
complications in 2.66% of them. The rate of device failure was 2%; therefore, the
unwanted outcomes requiring surgery were 4.66%, while 10% reported minor complication
rate.[19] Xu et al.,[20] in a meta-analysis study of 799 cases, reported no statistically significant difference
in major and minor complications between the two approaches, except for facial nerve
and chorda tympani injuries, favoring SMA.[20] Migirov et al.[21] operated 300 cases of CI and reported a rate of 3% major complications and 25.7%
minor complications.[21] Júnior et al.,[8] in a study of 250 cases, reported 13 cases (5.2%) of major postoperative complications
and 20 cases (8%) of minor complications.[8]
Kevin et al.[22] reported that the most common reasons for revision were device malfunction. They
found that 7.8% had device malfunction, out of 805 cases operated by both approaches.
Fayad et al.[23] reported 7% postoperative flap breakdowns with implant extrusion. Ajalloueyan et
al.[24] reported 1% of device malfunction in 262 patients that underwent CI.
Ajalloueyan et al.[24] reported that, out of the 262 patients on which the PT approach was used, two had
wound infection (0.8%) post CI[24]. Junior et al.[8] reported that 5 cases (2%) had wound infection post CI, out of 250 cases by PT approach.
Migirov et al.[21] reported that two cases had facial nerve paralysis (0.66%), out of 300 cases operated
by the PT approach. Using the same operative approach, Ajalloueyan et al.[24] reported that 1% of the 262 patients operated had facial nerve paralysis. Current
reports of facial nerve injury during facial recess surgery for CI show a consistent
rate of less than 1%, with several studies reporting a rate of 0.7%.[25]
[26]
[27]
[28]
[29] Nonetheless, none of these techniques led to facial nerve paralysis. The study showed
that the non-mastoidectomy approaches could avoid facial nerve and chorda tympani
injuries in patients with CI with low rates of other complications.
After analyzing the literature, we are convinced that non-mastoidectomy (SMA, tanscanal,
pericanal) approaches for CI are significantly faster, easier and safer in that which
pertains to facial nerve, when compared to the PT approach. In over one thousand published
cases of SMA, total incidence of complications were 11%. Furthermore, major complications
were detected in only 1.3% with transcanal, while a total of 9% reported no complications
with the permeatal approach. These complications rates are similar to those found
in the mastoidectomy approach, albeit with the additional advantage of shorter duration
of surgery, anesthesia, recovery and hospital stay.
This study also reveals that electrode fixation was not a matter of concern in the
techniques analyzed, given that the rate of electrode extrusion in over 1000 patients
was 0.79% in SMA and 1.88% for the transcanal approach. Junior et al.[8] reported one case (0.4%) of electrode exclusion out of 250 cases that underwent
CI by the PA approach, and Raghunandhan et al.[30] reported another single case (0.3%), out of 300 cases that underwent CI by the PT
approach.
Final Comments
Alternatively, non-mastoidectomy approaches are valid in cases where conventional
PT is difficult to perform. Signs that may indicate this method include narrow facial
recess, anteriorly located facial nerve, ossified cochlea (due to the promontory is
more exposed in details in non-mastoidectomy approaches better than classic approaches)
and severely-contracted mastoid. Therefore, the authors consider it helpful and believe
it should be mandatory that CI surgeons become familiar with these approaches.
