Hearing preservation cochlear implantation (CI) assumes atraumatic electrode insertion [
The aim of this study was to determine if cochlea metrics can be reliably obtained using routine imaging software and whether cochlear duct length and cochlear diameter were a factor in determining hearing outcomes for two different length electrodes.
The study was conducted at a tertiary adult implant center with ethics approval. A retrospective chart review of patients receiving a cochlear implant over a 5 year period was conducted. A total of 55 postlingual deafened adults were included. CI was offered after a failed trial of hearing aids and a best aided HINT sentence score in quiet of less than 60% [
Preoperative testing included a pure tone average (PTA) at 250Hz, 500Hz and 1000Hz, and word discrimination with CNC monosyllabic word testing in quiet at 60dB SPL [
Three surgeons performed surgery. A routine postauricular approach with facial nerve monitoring was used. A single dose of 8mg of dexamethasone was given on induction. A cortical mastoidectomy was performed followed by a posterior tympanotomy through the facial recess. The round window niche was lowered and after bone dust was removed and the round window membrane was opened and the electrode inserted gradually until full insertion or resistance. Insertion was performed using a combination of freehanded or instrument guided (using surgical claw, micro angled forceps or jeweller’s forceps). A small soft tissue plug was placed in the niche, with the remainder of the electrode coiled in the mastoid cavity before closure.
The Flex 31 electrode measures 31.5mm with 19 platinum electrode contacts spaced over 26.4mm. The diameter at the basal end is 1.3mm and the tip measures 0.5mm x 0.4mm. The Flex 28 electrode measures 28mm with 19 platinum electrode contacts spread over a shorter distance of 23.1mm. The diameter at the basal end is smaller at 0.8mm but the tip dimensions are the same as the longer electrode [
Only patients who had a preoperative temporal bone CT scan performed at our centre were included. The temporal bone CT scan was performed on a GE Lightspeed Plus 64 multi-slice CT scanner. The axial images were 0.625mm in thickness and the oblique-coronal plane images were 0.6mm in thickness. Image processing and measurements were performed on a GE AW-workstation release 4.4 running the Volume Viewer software version 8.3.65.
The cochlea was reformatted in the oblique coronal plane to obtain the entire basal turn in a single view. The straight measuring tool was then used again to measure the distance from the most lateral bony wall, through the modiolus, to the interface between air and soft tissue at the round window midpoint. This was termed the A-value measurement as per Escude (Figure
The A value measured in the oblique coronal plane (a). A straight line-measuring tool (b) is used to measure the outer wall cochlear duct length to 360 degrees. A side profile view (c) indicates the individual points used to calculate the length till 720 degrees.
The cochlea was then centered about its modiolus so that by scrolling superiorly the cochlear turns were gradually brought into view. A curved measuring tool was then used to measure the distance starting at the lateral wall of the round window (most proximal portion of the basal turn of the cochlea) and followed until 360 degrees of rotation was reached (Figure
Postoperatively, CNC word scores were measured at 6 months in all subjects. In the Flex 31 group, the HINT sentence test was performed at 6 months, and in the Flex 28 group the AZBio sentence test in quiet was also performed at 6 months. Postoperative plain X-rays of the skull were performed within 24 hours of after surgery using modified Stenver’s view. A senior radiologist blinded to the electrode type viewed the images. The degree of electrode insertion and the number of electrodes within the cochlear were reported [
There were 34 patients implanted with the Flex 31 and 21 patients implanted with the Flex 28 electrode. The mean age for both groups was 62 years and 63 years of age, respectively, with no significant difference the two groups. The mean age for the entire cohort of patients was 62 years (SD 12.8 years). There were 30 male and 25 female patients and 24 left and 31 right ears implanted. The etiologies of hearing loss were similar between the two groups with the majority of hearing loss due to idiopathic progressive sensorineural hearing loss (Table
Preoperative demographics. SD = standard deviation. P-value obtained using the Mann-Whitney U test and Chi square test.
Group | n | Mean age of implantation | Sex | Side of |
---|---|---|---|---|
Flex 31 | 34 | 62 (14.5) | 20:14 | 16 Left |
18 Right | ||||
| ||||
Flex 28 | 21 | 63 (9.8) | 10:11 | 8 Left |
13 Right | ||||
| ||||
Combined | 55 | 62 (12.8) | 30: 25 | 24 Left |
31 Right | ||||
| ||||
p-value | 0.81 | 0.58 | 0.58 |
The preoperative speech and PTA measures are shown in Table
Hearing and insertion outcome measures for each group. PTA = pure tone average at 250Hz, 500Hz and 1000Hz, HINT = hearing in noise test, and CNC = consonant nucleus consonant. P-value obtained using the Mann-Whitney U test, with significant values indicated in italic. Numbers in brackets indicate standard deviations.
Flex 31 | Flex 28 | p-value | |
---|---|---|---|
Preo op | |||
PTA (dB) | 69 (13.2) | 72 (8.9) | 0.555 |
Best aided CNC quiet (%) | 12.3 (15.8) | 30.9 (16.1) | |
HINT sentence quiet (%) | 36 (24.9) | 45.2 (22.2) | 0.218 |
AZBIO sentence quiet (%) | n/a | 33.3 (20.6) | |
6 months post op | |||
Best aided CNC quiet (%) | 52.6 (22.8) | 59.7 (17.9) | 0.399 |
HINT sentence quiet (%) | 83.7 (21.8) | n/a | |
AZBIO sentence quiet (%) | n/a | 63.3 (19.99) | |
CNC score shift (%) | 40.7 (24.1) | 28.9 (17.4) | 0.055 |
Insertion outcomes | |||
(1) Degrees of electrode insertion on plain X-ray | 489.4 degrees | 525 degrees | 0.165 |
(2) Number of channels inserted on plain X-ray (out of 12) | 10.85 (1.08) | 11.33 (0.73) | 0.12 |
Correlation | | | |
between (1) and (2) | (p<0.01) | (p<0.01) |
The postoperative CNC word scores at 6 months were comparable between the two groups at 52.6% (Flex 31) and 59.7% (Flex 28) (Table
The results of the different cochlear metrics are shown in Table
Cochlear metrics for all subjects. There is no statistically significant difference in all measurements between subjects receiving either electrode.
Minimum | Maximum | Mean | Standard Deviation | ||
---|---|---|---|---|---|
A-vaue | Flex 31 | 8.1 | 9.8 | 8.94 | 1.63 |
Flex 28 | 8.1 | 9.5 | 8.87 | 0.38 | |
Combined | 8.91 | 0.37 | |||
| |||||
Outer wall length to 360 degrees | Flex 31 | 29.8 | 35.9 | 32.52 | 1.41 |
Flex 28 | 27.8 | 34.7 | 31.91 | 1.63 | |
Combined | 32.29 | 1.51 | |||
| |||||
Outer wall length to 720 degrees | Flex 31 | 19.6 | 23.5 | 21.4 | 0.94 |
Flex 28 | 18.3 | 22.6 | 21.1 | 1.09 | |
Combined | 21.3 | 1 |
A histogram showing the normal distribution of the outer wall CDL to 720 degrees in all 55 patients.
In the Flex 28 group, the A-value was correlated with the postoperative CNC word score (R=0.64) and the AZBIO score(R=0.46). In other words, when the shorter electrode was used, a larger basal diameter was associated with better speech discrimination (Table
A-value correlations with
Flex 31 | Flex 28 | ||||
---|---|---|---|---|---|
CNC (%) | HINT (%) | CNC (%) | AZBIO (%) | Electrodes inserted | |
Ac | 0.36 | 0.34 | | | 0.19 |
Outer wall length 7200 | 0.19 | 0.19 | | | 0.27 |
Degree of insertion | 0.23 | 0.22 | -0.08 | -0.09 | |
Electrodes inserted on X-ray | 0.26 | 0.24 | 0.054 | -0.08 |
A plot showing a statistically significant correlation between the Ac value and the AZBio sentence score (a) and CNC word score (b) in the Flex 28 group as well as the correlation between the outer wall cochlear duct length and the AZBio sentence score (c) and CNC word score (d) at 6 months.
In both groups, the degree of insertion or the number of channels inserted did not correlate with speech outcomes at 6 months. The degree of electrode insertion and number of electrodes inserted was higher in the Flex 28 group (525 degrees and 11.3 channels) compared to the Flex 31 group (489.4 degrees and 10.9 channels) although this was not significantly different. For both groups, the number of channels inserted on plain X-ray was positively correlated with the degrees of electrode insertion, with a stronger correlation for the Flex 28 group (R=0.71 vs. R=0.589). In other words, the greater the degree of electrode insertion the greater the number of channels visualized to have been inserted on plain X-ray. In the Flex 28 group and the Flex 31 group, neither the degree of insertion nor the number of channels inserted was correlated with any cochlear measurements.
Previous cadaveric studies have shown that CDLs and the number of cochlear turns varies between individuals [
The technique used to measure the CDL and the A-value was chosen for its ready availability. All our cochlear implant candidates now have standardized temporal bone CT scans as part of their workup and no specialized software is required to process the images unlike other automated techniques [
The plain postoperative X-ray was used to determine angular insertion [
When comparing CDLs across different studies, a standardized method needs to be utilized in order to make meaningful comparisons, a need which has been echoed by other authors [
In comparison to these studies, our study measured the outer wall CDL to 720 degrees across 54 patients and found a mean length of 32.29mm. The human cochlea may vary from 774 to 1037 degrees, with a mean of 929 degrees [
Our mean basal diameter value of 8.91mm is slightly longer than Martinez-Monedero’s value of 8.39mm (SD 0.76) and Ketten’s value of 7.91mm (although the central fluid space was used as the reference in the latter) but slightly shorter than Escude’s value of 9.23mm (SD 0.53) and Connor’s value of 9.36mm (SD 0.31) [
In our study, the basal turn represented 65.9% of the CDL to 720 degrees. This is comparable to Hardy’s figure of 57.9%, Escude’s figure of 59%, and Erixon and Rask-Andersen’s figure of 53%
Our insertion angles with the Flex 28 are similar to that obtained by Franke-Trieger (Franke-Trieger et al., 2013). However, our results indicate a lower insertion angle for the Flex 31 electrode compared to previously published figures which also vary considerably. Franke-Trieger in the same 10 adult temporal bones found a mean insertion angle of 673 degrees. However, in Trieger’s paper, complete insertion was not achievable in all subjects with the 31mm electrode [
We found only two studies with lower insertion angles than our study for the Flex 31 electrode. The first was a study by Radeloff using a Med-El Combi 40+ electrode (31.5mm) via a cochleostomy in a 28 temporal bones [
Our study found no correlations between cochlear size and the degree of insertion. This finding is unlike previously reported by two studies [
Our study found no statistical correlation between the degree of insertion and the postoperative outcomes at 6 months with both the length electrodes. This is consistent with several previous studies [
There are limitations of comparing the 31mm with the 28mm electrode for different sized cochlea. Ideally one should compare the same sized cochlea and compare the performance between the two different lengths electrodes that are inserted in the same manner. However, practically, it would be difficult to find two exactly similar sized cochleae in all dimensions including length, height, and width as well as with the same level of hearing loss and spiral ganglion distribution. There are also other factors which cannot be controlled for such as incomplete insertion. Surrogates have to therefore be utilised and in this case we have chosen to choose age, hearing loss, and a well-established cochlear metric.
When each individual electrode group was analysed separately, larger cochlear sizes were associated with better speech performance for the Flex 28 group. When the Flex 31 electrode was used, there did not appear to be any correlations between cochlea size and speech performance. What is interesting is that this association is electrode specific and needs to be reconciled with the observation that insertion depth and speech perception are not well correlated. The reasons for this are unclear and need to be explored in future studies. A shorter electrode may be less traumatic in a relatively longer cochlea, leading to a less traumatic insertion. A larger cochlea may also have a different distribution of spiral ganglion neurons which may be more conducive to stimulation by a shorter electrode. A smaller electrode in a larger cochlea may also lead to a lower risk of postoperative electrode migration, a factor poorly studied in the literature. As this is a retrospective study, this finding does not necessarily indicate that one should choose a shorter electrode for a relatively longer cochlea. What this does indicate is that cochlear size may be an important factor in determining CI outcomes. Attempts to predict the appropriate electrode length relative to cochlear size need to be examined to determine the effectiveness of such a technique [
A study by Johnston et al. is very similar to ours and compared retrospectively the outcomes of Flex 28 and Flex 31 electrodes [
Cochlear metrics can be measured using routinely available radiological software using the preoperative temporal bone CT scan. The size of the cochlea appears to be an important factor which may affect CI outcomes for certain electrode lengths. This factor should be considered in future strategies for electrode selection.
The data used to support the findings of this study are restricted by the Sunnybrook Human Ethics Research Committee in order to protect patient privacy. Data are available from Dr. Jafri Kuthubutheen for researchers who meet the criteria for access to confidential data.
An earlier version of this study has been presented as a poster at the Triological Society 2014 Combined Sections Meeting.
The authors declare that there are no conflicts of interest regarding the publication of this paper.
The authors acknowledge the Unrestricted Departmental Research Grant, Sunnybrook Health Sciences Centre, Department of Otolaryngology Head and Neck Surgery.