Primary Stapedectomy: A Retrospective Review of Surgical Results Using a Cold Steel Technique and Pre-Operative CT Correlates

Abstract

Importance

Stapedectomy is an effective treatment option for patients with symptomatic otosclerosis. A common surgical technique today is laser stapedotomy, although the equipment required adds a significant cost to the procedure and may not be available in certain, particularly remote, centers. This study aimed to evaluate cold steel stapedectomy in a modern otology practice as a cost-effective and versatile alternative.

Objective

The primary objective was to evaluate the efficacy and safety of cold steel stapedectomy. The secondary objectives included characterizing the prevalence of relevant computed tomography (CT) findings and evaluating other factors such as piston size.

Design

A retrospective case series from April 2006 to November 2021.

Setting

A single tertiary care hospital in Montreal, QC, Canada.

Participants

Patients with suspected otosclerosis and no prior stapes surgery.

Intervention

Cold steel primary stapedectomy.

Main Outcome Measures

In primary analysis, change in pure tone thresholds, pure tone average (PTA), and complication rates were obtained. Secondary analysis measures included rates of relevant CT findings and change in PTA with piston size.

Results

A total of 302 cases were included in the analysis. The average change in air conduction PTA was a decrease of 27 dB, with 91% (276/302) having over 10 dB of improvement. The air-bone gap was closed to within 10 dB in 82% (248/302) of cases and within 20 dB in 96% (290/302) of cases. Overclosure occurred in 19% of cases, while only 4% resulted in worsened PTA bone conduction thresholds. In secondary analysis, 74% of CT scans demonstrated radiologic otosclerosis. Other findings included suspected superior semicircular canal dehiscence in 1.8%.

Conclusion

Cold steel stapedectomy was demonstrated to be a safe and effective technique, with audiometric results comparable to laser and drill studies in recent literature. It should be considered as a cost-effective and global health accessible alternative. Additionally, CT scans can provide valuable information in the pre-operative workup.

Graphical abstract.

Background

Otosclerosis is a fibrous osteodystrophy of the otic capsule, characterized by abnormal bone remodeling and fixation of the stapes footplate. ¹ Clinically, it manifests primarily as a hereditary cause of conductive hearing loss, with sensorineural and mixed hearing losses seen frequently as well. Although previous reports have suggested that most patients with histologic evidence of otosclerosis are asymptomatic, ¹ symptomatic cases are found to have a progressively-worsening hearing loss in one or both ears that can reach 50 to 60 dB of loss. Currently, there are 3 commonly-accepted management options available: observation, hearing amplification (ie, hearing aids), or stapes surgery.

Surgical options to treat otosclerosis include stapedectomy or stapedotomy. In the classic stapedectomy, part or all the stapes footplate is removed using cold steel instruments to place a piston prosthesis through the oval window. ² In stapedotomy, a small hole is made with a laser or drill just large enough to fit a similar prosthesis. The laser has become a popular option in modern practice as it offers a minimally-invasive method of dissecting the stapes and opening the inner ear, but its major drawback is the high equipment cost, making it not available at all centers (particularly in low-resource settings). A series of studies have compared the 2 procedures, ³ with mixed results found across several metrics. For pure tone audiometry, multiple studies suggested stapedotomy had superior results, most notably at high frequency (8 kHz).^4-6 On the other hand, 1 study revealed superiority in stapedectomy across all frequencies, ⁷ and others found no difference.^8-10 Similarly, for air-bone gap closure, some studies favored stapedotomy,^6,10 but others suggested either stapedectomy as superior^7,8 or no difference. ⁴

Imaging in the preoperative workup of otosclerosis has not been uniformly accepted in the current standard of care. Although physical findings and audiometry can most often provide reasonable certainty in the diagnosis, high resolution computed tomography (HRCT) imaging can confirm equivocal cases or can reveal other important prognostic signs such as retro-fenestral disease. HRCT can also evaluate for confounding pathologies such as ossicular chain disruption and middle ear masses or can reveal surgical concerns such as facial nerve dehiscence or a high riding jugular bulb. A series of previous studies evaluated the diagnostic accuracy of computed tomography (CT), ¹¹ which demonstrated sensitivity ranging from 61% to 95% and specificity 75% to 100%. Other studies have compared CT findings with audiometry results,^12,13 confirming the hypothesis that more extensive radiologic disease correlates with worse pre- and postoperative hearing. Still, many otologists choose not to scan their patients preoperatively.

Given that a common opinion in the current otologic community is that laser is the superior method, that many of the previous studies could not conclusively prove one over the other, and that most studies evaluating cold steel technique are several decades old, the primary objective of this study was to evaluate the surgical results of a single surgeon performing cold steel stapedectomy at a modern tertiary care institution. These results were compared with those of previous literature on laser stapedotomy. The second objective was to evaluate the merits of preoperative CT scans, more specifically examining the rate of confounding pathologies and surgically-relevant findings.

Methods

Study Design

A retrospective case series was performed evaluating patients who underwent primary stapedectomy using a cold steel technique by a single surgeon (J.M.R.) at a tertiary care institution (Jewish General Hospital, Montreal, Quebec, Canada) from January 2006 to November 2021. A subset of these patients who had available CT temporal bone scans were included in a secondary analysis. The study was approved by the research ethics committee at the Jewish General Hospital (2022-3129).

Patient Selection

Included patients were ≥18 years old with a diagnosis of otosclerosis and had a cold steel primary stapedectomy performed by a single surgeon. Patients were excluded if pre- or postoperative audiometry was not available at the time of data collection. For the secondary analysis of imaging data, patients were included if they had a high-resolution CT scan of the temporal bone that was previously analyzed by a subspecialized head and neck radiologist.

Surgical Technique

The approach begins in standard fashion by raising a posteriorly-based tympanomeatal flap and accessing the middle ear below the annulus. Bone of the scutum is carefully curetted away until the stapes and stapedial tendon are in clear view. The malleus is palpated to ensure there is no incudomalleal fixation, and the stapes is confirmed to be rigidly fixated. The incudostapedial joint is then divided with a joint knife, the stapedial tendon is cut with Bellucci scissors, and the stapes suprastructure is downfractured using a curved pick. The suprastructure is removed exposing the underlying footplate, and the distance from footplate to incus is measured with a depth gauge. A safety hole is then made in the center of the footplate with a straight pick to ensure there is no gusher of perilymphatic fluid. This is then postage-stamped in the mid-portion of the footplate with the straight pick, and the posterior portion of the footplate is then carefully lifted from its anterior edge using a 45° hook and removed. A Schuknecht Fluoroplastic stainless wire piston prosthesis (Olympus Gyrus^®) is placed in the stapedotomy and crimped on the long process of the incus manually, and a small fat graft taken from the ear lobe is placed in the oval window niche around the piston to seal the stapedotomy. Finally, the ear is closed by laying the tympanomeatal flap back onto the posterior canal wall and the external ear is packed with silk strips and cotton fashioned into a rosebud dressing.

Data Collection

The following data were retrieved for each patient: demographics, preoperative audiometry (including pure tone and speech), relevant operative findings, prosthesis size, intraoperative and postoperative complications, and postoperative audiometry. For the subset with relevant CT scans, the presence of radiologic otosclerosis (fenestral and retrofenestral) and other surgically-relevant findings (high riding jugular bulb, dehiscent facial nerve, etc) were collected.

Statistical Analysis

All statistical analysis was performed in R v4.0.2 (R Foundation for Statistical Computing, Vienna, Austria). For the primary surgical analysis, a comparison between mean pre-operative to post-operative air conduction thresholds, bone conduction thresholds, and air-bone gaps at individual frequencies was performed using a Student’s t-test. Pure tone averages were calculated using 0.5, 1, 2, and 4 kHz for air and bone conduction. Prosthesis diameter and length were similarly compared using Student’s t-test of hearing change (mean air conduction pure tone average change). For the CT analysis, the rate of each relevant finding was reported as a simple percentage, and secondary analysis compared hearing improvement based on the presence or lack of radiologic otosclerosis by t-test.

Results

Surgical Results

A total of 302 patients were included in the analysis. Baseline characteristics, including demographics and prosthesis sizes, are listed in Table 1. The average age at the time of operation was 49.3 years (σ = 11.1), and the patients were mostly female (63% vs 37% male). Prosthesis lengths were chosen as per intraoperative measurement and were mostly 4.25 mm (84%), followed by 4.5 mm (15%), 4.0 mm (0.7%), and 4.75 mm (0.3%). The primary prosthesis diameter used was 0.6 mm (85%), while the 0.8 mm was used in cases where the stapedotomy was larger than the typical posterior fenestra technique (ie, removal of most or the entire footplate).

Table 1.

Baseline Characteristics (n = 302).

Age (mean)	49.3
Gender
Male	111
Female	191
Laterality
Right ear	169
Left ear	133
Prosthesis length (mm)
4.75	1
4.5	45
4.25	253
4.0	2
Prosthesis diameter (mm)
0.8	42
0.6	258

The change in air conduction thresholds from pre- to postoperation can be seen in Figure 1. To mimic the appearance of a standard audiogram, mean values are reported at each of the fundamental frequencies with the dependent axis in reverse order. Statistically-significant improvement was demonstrated at all frequencies except at 8000 Hz, with improvements ranging from over 30 dB in the lower frequencies down to approximately 15 dB at 4000 Hz. The change in bone conduction thresholds was demonstrated similarly in Figure 2. Although stapedectomy is typically thought to address conductive hearing loss in otosclerosis, there was a small but statistically-significant improvement in the bone conduction thresholds from 500 Hz to 2000 Hz. There was also a very small but significant worsening of the threshold at 4000 Hz. There was an inadequate amount of 8000 Hz bone conduction data to be included.

Figure 1.

Mean air conduction pure tone thresholds pre- and post-stapedectomy. NS, not significant.

Shown as mean ± SEM.

*P < .05, **P < .01, ***P < .001.

Figure 2.

Mean bone conduction pure tone thresholds. NS, not significant.

Shown as mean ± SEM.

*P < .05, **P < .01, ***P < .001.

The distribution of pre- and postoperative air-bone gap by frequency can be seen in Figure 3. There was a statistically-significant drop in the air-bone gap across all frequencies. The most pronounced effect was in the lower frequencies, in keeping with the air conduction improvement pattern seen in Figure 1. The improvement was lowest at 2000 Hz, likely due to the preoperative bone conduction deficit commonly seen in otosclerosis (the Carhart notch) that was shown to improve with treatment (see Figure 2). The overall rate of air-bone gap closure (<10 dB) was 82% (248/302).

Figure 3.

Air-bone gap pre- and post-stapedectomy. NS, not significant.

*P < .05, **P < .01, ***P < .001

A summary of important audiometric results can be seen in Table 2. The mean change in pure tone average air conduction threshold was an improvement of 27.1 dB, with 91% having an improvement of at least 10 dB and 44% greater than 30 dB. For bone conduction, the mean change was an improvement of 3.5 dB, with an overclosure of greater than 10 dB in 19% of patients and worsened thresholds of 10 dB or more in only 4%. A simple prosthesis size analysis compared length and diameter of pistons based on the improvement of the air conduction pure tone average. The pistons inserted in most of these cases had a length of 4.25 or 4.5 mm, and their comparison revealed no difference in audiometric result (P = .67). Other piston lengths were omitted in this analysis (n = 3). For piston diameter, there was a trend toward greater improvement with the 0.6 mm piston, but this did not demonstrate statistical significance (P = .11).

Table 2.

Overall Results.

Audiometric
Pure tone average change
Air conduction (mean)	−27.1 dB (SD = 13.7 dB)
Improvement >10 dB	91% (276/302)
>20 dB	77% (232/302)
>30 dB	44% (132/302)
Bone conduction	−3.5 dB (SD = 10.6 dB)
Overclosure (>10 dB increase)	19% (56/302)
Significantly worsened threshold (>10 dB decrease)	4% (11/302)
Prosthesis size analysis a
Length
4.25 mm	−27.1 dB	P = .67
4.5 mm	−27.9 dB
Diameter
0.6 mm	−27.9 dB	P = .11
0.8 mm	−22.9 dB
Surgical
Intraoperative—sclerosis
Footplate sclerosis—resulting in total stapedectomy	7.9% (24/302)
Severe footplate sclerosis—requiring drill	1.7% (5/302)
Complications
Vertigo and/or dizzy complaints b	10.3% (31/302)
Taste complaints b	2.0% (6/302)
Perforation b	0.7% (2/302)
Profound hearing loss	1.0% (3/302)

^aChange in mean air conduction threshold.

^bBased on early postoperative assessment (none noted at later visits).

In terms of surgical results, the presence of significant sclerosis of the footplate was recorded and can also be seen in Table 2. In 7.9% of cases, there was significant sclerosis of the footplate, rendering it too rigid for the posterior fenestra technique and the result was the footplate removed in its entirety with manipulation. In 1.7% (5/302) of cases, the footplate was severely sclerotic, requiring the use of a skeeter drill to create a stapedotomy. With respect to complications, the most common was vertigo or dizziness complaints, which occurred in 10.3% of cases. These were all temporary, with no patients noted to have lasting dizziness on later postoperative follow-ups. Taste complaints were noted in 2% and perforation noted at first follow-up visit in 0.7% (2/302). The most feared complication, profound sensorineural hearing loss, was seen in 1% of cases (3/302).

Radiology Correlation

A total of 112 patient CT scans were included in this analysis. A sample of relevant radiologic findings can be seen in Figure 4, demonstrating classic antefenestral disease, retrofenestral disease, a high riding jugular bulb, and radiologic superior semicircular canal dehiscence. The rates of all relevant findings are summarized in Table 3. Clinically-relevant findings noted were ipsilateral antefenestral, bilateral, retrofenestral disease, and superior semicircular canal dehiscence, with rates of 74.1%, 53.6%, 7.1%, and 1.8%, respectively. Note that the 2 patients with suspected superior canal dehiscence did not complain of the classic third-window symptoms (autophony, Tullio phenomenon, etc) in this series. Surgically-relevant findings noted were a high riding jugular bulb or a suspected dehiscent facial nerve in 4.5% and 1.8%, respectively. A brief analysis comparing radiologically-positive otosclerosis cases to radiologically negative based on audiometric improvement yielded no statistically-significant difference (P = .32), implying that the presence of otosclerosis on scan did not predict a better or worse surgical outcome.

Figure 4.

Sample of radiologic findings noted in patients with otosclerosis. (A) Classic antefenestral disease (*). (B) Retrofenestral disease. (C) High riding jugular bulb (*). (D) Radiologic superior semicircular canal dehiscence (*).

Table 3.

Radiologic Results (n = 112).

Findings
Ipsilateral antefenestral disease	74.1% (83/112)
Bilateral antefenestral disease	53.6% (60/112)
Retrofenestral disease	7.1% (8/112)
High riding jugular bulb	4.5% (5/112)
Suspected superior canal dehiscence	1.8% (2/112)
Suspected dehiscent facial nerve	1.8% (2/112)
Audiometric comparison (∆AC-PTA)
Positive radiologic otosclerosis	−27.3 dB	P = .32
Negative (no radiologic otosclerosis)	−30.1 dB

Abbreviation: AC-PTA, air conduction-pure tone average.

Discussion

The original stapedectomy procedure described by Shea in 1958 ¹⁴ involved a complete removal of the stapes with cold steel instruments and the insertion of a biocompatible prosthesis connecting incus to the oval window. Since then, multiple advancements have been proposed and adopted, most notably the use of either a laser or micro-drill for a minimally-invasive stapedotomy. And although there is a commonly-held belief in the otologic community that these new approaches reduce trauma to the inner ear and result in less sensorineural hearing loss, less dizziness complaints, and improved hearing outcomes, previous comparative study results were equivocal. This study aimed to evaluate use of a more traditional cold steel approach in a modern surgical practice.

Across many audiologic metrics, this cold steel large fenestra technique was demonstrated to produce overall positive results. For air conduction, this study revealed a mean pure tone average improvement of 27.1 dB and an overall air-bone gap closure rate of 82%. Multiple large previous studies have evaluated for small fenestra stapedotomy, with heterogeneous reporting of results. For pure tone thresholds, House et al ¹⁰ (based on a combination of laser and drill techniques) reported mean pre- and postoperative air conduction-pure tone average (AC-PTA) of 55.9 and 33.5 dB, respectively, implying a mean improvement of approximately 22.4 dB. For air-bone gap, Sedwick et al ⁸ (drill and laser stapedotomy) reported closure rates of 78.2%, and Colletti et al ⁴ (drill) reported 86.5% for 0.5 to 2 kHz and 80.7% for 4 kHz. One systematic review specific to laser stapedotomy analyzed a collection of smaller studies, ¹⁵ where air-bone gap closure rates ranged from 72.2% (39/54) to 93% (93/100), collectively resulting in 84.5% (539/638). In terms of bone conduction, the analysis above demonstrated a slight improvement in pure tone average of 3.5 dB. In previous literature on stapedotomy, House et al ¹⁰ reported a bone conduction-pure tone average (BC-PTA) improvement of 4.5 dB and Kos et al, ¹⁶ an improvement of 3.0 dB. Ultimately, although heterogeneity of reporting makes it challenging to make direct comparisons between studies, there are clear audiometric similarities in the results of the cold steel technique with those of stapedotomy.

Some complications were noted in this study. Significant sensorineural hearing loss (decreased BC-PTA >10 dB) was observed in 4% of cases. Previous literature on stapedotomy described rates ranging from 6.9% (minimum 10 dB PTA loss), ¹⁷ to 0.62%, ⁷ with other studies showing 5.9%, ¹⁰ 3.2%, ⁸ and 0.69%. ⁵ The most feared complication, severe-to-profound sensorineural hearing loss (SNHL), was observed here in 3 patients (1%), while prior studies of stapedotomy (although not consistently reported) demonstrated rates from 0.9% ¹⁶ to no cases.^7,10 Further study would be required to determine whether there is a statistically-significant difference of profound loss. More common complications included dizziness and/or vertigo complaints (10.2%) and taste changes (2%). All cases of these symptoms were transient, with no residual issues noted on further follow-up. Sedwick et al ⁸ reported transient vertigo in 23.3% and taste disturbance in 16.2%, but they did not distinguish based on stapedectomy versus stapedotomy, and Parilla et al ¹⁸ reported transient vertigo in 3% of patients who received laser stapedotomy. Overall, as with the pure tone audiometry, a lack of uniform reporting makes it difficult to compare across many studies, but the complication rates reported here for cold steel stapedectomy fall within a similar range as other reports of stapedotomy.

A brief analysis of prosthesis size was conducted. In this sample of patients, the prosthesis lengths inserted in the overwhelming majority of cases were 4.25 or 4.5 mm, and diameters of 0.6 or 0.8 mm. Statistically, comparing the AC-PTA between the 2 lengths yielded no difference, supporting the previously-described finding that the best prosthesis length is the 1 measured (ie, not 1 optimal size). ¹⁹ Similarly, diameter did not reveal a significant difference, implying that the use of a larger diameter piston for a larger fenestra did not compromise results. Another brief secondary analysis evaluated the intraoperative sclerosis in terms of the surgical management required. In 7.9% of cases (24/302), the posterior fenestra technique was not possible due to footplate sclerosis, and instead, removing the whole footplate was the only option with cold steel instruments. In 1.7% of cases, the sclerosis was severe enough that a drill was required to access the inner ear. This is an important consideration with a cold steel technique, as a drill would need to be available for this low but real possibility or risk having a rare incomplete surgery. The same would apply to laser stapedotomy, as the laser cannot be used alone to get through a severely-thickened footplate.

Imaging in otosclerosis is a controversial topic. Although previous systematic review has demonstrated that high-resolution CT scans of the temporal bone provide some diagnostic value, ¹¹ the lack of sensitivity or specificity of radiologic antefenestral demineralization in predicting clinically-relevant disease makes it a weak indication on its own. Instead, many surgeons tend to rely on scans for 3 other reasons. The first and likely most important is confounding pathologies. This study demonstrated suspected superior semicircular dehiscence in 2 cases (1.8%), a finding that requires careful preoperative assessment and counseling as there is the risk of unmasking third window symptoms. Another important finding would be an ossicular discontinuity or malformation, which was not demonstrated here, but this study design would be unlikely to do so (see limitations below). The second reason to get a CT is for surgically-important findings. This study demonstrated a high riding jugular bulb in 5 cases (4.5%) and suspected dehiscent facial nerve in 2 (1.8%), and no injuries to either were noted in the operative reports or postoperative assessments. One can also identify other abnormalities, such as an enlarged vestibular aqueduct or other congenital temporal bone finding, which might be associated with an increased risk of perilymphatic gusher or sensorineural hearing loss. The third reason would be for prognostically-relevant findings. This study demonstrated contralateral radiologic disease in 53.6%, and retrofenestral disease in 7.1%. Marx et al ¹² previously compared audiometric findings for CT-positive disease over a 3 years period, showing that the presence of radiologic disease in the contralateral non-operated ear led to worsening of both its air and bone thresholds over time. They did not, however, provide an analysis of the long-term changes in operated ears with retrofenestral disease. Regardless, these findings are helpful to discuss with patients preoperatively to manage expectations over time.

There were several limitations in this study. First, it was retrospective and is subject to the inherent biases of retrospective review. The procedures were performed by a single surgeon at a single hospital, and there was insufficient late audiometric data for long-term analysis. Postoperative audiograms were not consistently performed at the same center, and their timing pre- and postoperatively was not consistent. Some patients had to be excluded because their follow-up was at a private clinic with private audiology (unavailable for analysis). Similarly, many CT scans were excluded because they were performed externally, read by general radiologists and often without high-definition temporal bone acquisitions. CT scans were not retrospectively re-reviewed by a radiologist for this study, and some subtle findings may have been missed or not reported. Importantly, this analysis was exclusively on operative cases and therefore does not report whether there were patients who did not receive an operation because of a CT scan finding, and what that finding was.

Conclusions

This study demonstrated that stapedectomy with cold steel instruments is safe and effective, with audiometric and complication results comparable to previous literature on laser and drill stapedotomy. It should be considered a cost-effective, versatile option that is global health accessible. Performing preoperative CT scans, although still not clearly a necessity in otosclerosis, does provide valuable information in ruling out confounding pathology, assessing for intraoperative hazards and providing prognostic information for patients.