Reconstructive ossiculoplasty options in primary cholesteatoma surgeries with an intact stapes

Primary acquired cholesteatomas in children and adults might result in conductive hearing loss due to ossicular chain involvement. Erosion of the long process of the incus is most frequently observed [1]. The integrity and mobility of the stapes cannot be determined preoperatively, since even high-resolution CT or conebeam CT scans are often difficult to interpret due to obliteration of the oval window niche by cholesteatoma or granulation tissue [2]. To find an intact and mobile stapes offers ideal reconstructive options, including incus interposition with autologous or titan incus, stapes augmentation with autologous incus, cartilage or partial ossicular replacement prosthesis (PORP), and a type III tympanoplasty in case of a lowered tympanic membrane, mainly in open mastoido-epitympanectomy (oMET).

There are two main debates in hearing function after cholesteatoma surgery currently discussed: do open cavity surgeries provide less good hearing outcomes compared with closed cavity techniques and should ossiculoplasties be performed immediately or staged? Regarding the first, poor middle ear ventilation and more extensive preoperative pathology might impact postoperative hearing in open cavity surgery [3]. However, studies do not unequivocally report worse hearing outcomes after oMET compared to closed cavity surgery, even without considering the above-mentioned factors necessitating oMET surgery. Regarding the decision about concurrent or staged ossicular chain reconstruction in cholesteatoma surgery, a shift has been observed in the last decade. Previously, staging cholesteatoma surgeries was also intended to exclude residual disease at the second stage surgery. Nowadays, observation by non-EPI-diffusion MRI scans in closed cavity and otoscopically in open cavity surgeries is possible. Since more than 12 years, ossicular reconstruction at first surgery is attempted at our center. This is increasingly supported by literature results showing that staged cholesteatoma surgery seems only advantageous in patients with most severe disease and uncertainty about the complete removal of cholesteatoma [4, 5].

The aim of this study was the evaluation of hearing outcome in primary one-stage cholesteatoma surgery for children and adults in the presence of an intact and mobile stapes, a frequent finding at surgery. By including open and closed surgical techniques, we further intend to compare these, considering eventual modifiers such as the ossicular reconstruction method and the pneumatization/ventilation status of the mastoid, which are included in the ChOLE classification [6].

In cholesteatoma patients, different outcome parameters should be considered, such as recidivism (residual and recurrent cholesteatoma), surgical complications, and hearing. In the current retrospective study, we focused on hearing outcome after cholesteatoma surgery, namely patients with ossicular involvement but an intact and mobile stapes. All included patients underwent primary one-stage ossicular reconstruction. Since the availability of non-EPI-diffusion MRI scans, we do not routinely perform staged surgery anymore in closed cavity surgeries. In this way, we avoid increased hearing loss between the two surgeries and avoid unnecessary second look surgeries. In open cavities, there is no need for MRI follow-up since recidivism is observed clinically. One-stage hearing reconstruction seems to be supported by the rather low number of recidivism (9.5%, including 1.6% residual disease, 5.5% true recurrence, and 2.4% external ear canal or tympanic membrane inclusion cyst). Our results show a mean postoperative ABG of 20 dB (average between 500 and 4000 Hz), slightly better than preoperatively, and no significant difference in ABG after surgery. This confirms the clinical observation that cholesteatoma preoperatively still transmits sound to the intact stapes in cases of incus erosion.

Several parameters have been proposed to impact hearing outcome after cholesteatoma surgery. One of the main controversies is whether open versus closed surgical techniques provide better hearing. We found only a small but significantly worse postoperative ABG in oMET surgeries (23 dB) compared to closed cavity surgeries (19 dB). A higher number of type III tympanoplasties, which showed worse hearing outcome compared to the other reconstruction techniques, in oMET surgeries mainly explains this difference. A larger but not significant ABG in oMET surgery with incus interposition (24 dB) versus stapes augmentation (19 dB) was also noted. Hearing in patients who underwent oMET surgery with stapes augmentation proved in the same range as incus interposition or stapes augmentation in closed cavity surgery though. It is hard to compare open versus closed cavity surgeries in view of hearing outcome, as pre-existent factors influencing postoperative hearing have been detected. Examples are poor middle ear and mastoid ventilation and more advanced disease, which dominate oMET surgeries [3]. We believe a small and shallow middle ear cavity is a negative prognostic factor, as shown in type III tympanoplasties independent of an open or closed surgical technique. Middle ear cavity volume, mastoid pneumatization, the size and state of the antrum, and mucosal state are interrelated [8] and are approximated by the E status of the ChOLE classification [7]. The E status in our study was related in a borderline significant way to hearing outcome, with a better pneumatization/ventilation status associated to a lower postoperative ABG. The pneumatization/ventilation status might influence the choice of surgical technique, which in turn might influence the choice of ossiculoplasty technique, making a formal comparison difficult in absence of surgical technique randomization.

The presence or absence of the malleus handle, rather than the stapes suprastructure, was shown as a significant predictor of postoperative hearing outcome in a meta-analysis of patients with cholesteatoma or non-cholesteatoma chronic otitis media [9], which confirmed earlier observations of Dornhoffer et al. [10]. In our series, no significant difference between incus interposition and stapes augmentation could be observed. Therefore we advise stapes augmentation in case of a too far anteriorly positioned malleus handle.

When comparing ossicular reconstruction material within the groups, the better outcome in stapes augmentation with cartilage (16 dB) compared to autologous incus (23 dB) is apparent, although not statistically significant due to a lower number of patients. Ayache et al. thoroughly reviewed the procedure of cartilage stapes augmentation and compared with other autologous and synthetic materials [11]. They also concluded that cartilage stapes augmentation is a safe and effective procedure with no additional cost, and with hearing results at least as good as with other materials. The difference in incus interposition material, autologous versus titan incus, is less pronounced (18 dB versus 20 dB). A further benefit of cartilage or autologous bone is the fact that no extrusion can be observed.

Another interesting finding is the improvement in ABG after ossicular reconstruction surgery for 500, 1000 and 2000 Hz, but not for 3000 and 4000 Hz. At initiation, we decided to report the PTA of 500–4000 Hz as it covers the human speech recognition frequency range, although many other studies only report the PTA of 500–2000 Hz. The results of the latter are more beneficial, also in our series, but render comparison of absolute values among studies more difficult. We previously reported a postoperative increase in high-frequency ABG, mainly in type III tympanoplasty [12]. Here, a stable ABG at 3 kHz and increase at 4 kHz was noted in both open and closed cavity ossicular reconstructions. It is widely known that, preoperatively, cholesteatoma or granulation tissue might transmit sound, even in a discontinued ossicular chain [13]. Based on our results, ossicular reconstruction seems more effective in transmitting sound at the lower frequencies, but not at 4 kHz. In a temporal bone study with incus removal but intact malleus and stapes, several ossicular reconstruction methods were compared [14]. It was found that the optimal reconstruction is the one connecting stapes head to both the malleus handle and tympanic membrane, although poorer sound transmission in the higher frequencies was found for all conditions. However, gluing the contact between stapes head and prosthesis resulted in better high-frequency transmission [14]. It can thus be hypothesized that the higher ABG at 4 kHz after ossicular reconstruction might be related to a rather loose connection between stapes head and the ossicular reconstruction material. Therefore one might have to come up with a new technique of firmer connection towards the stapes head.

The strengths of this study are the consistency of surgical technique among the two surgeons, the availability of complete pure tone audiometry results and the reporting of 4 kHz enabling us to detect additional findings relevant for the patient. Weaknesses are the retrospective nature of the analysis despite prospective acquisition of data, and the lack of speech discrimination results due to inconsistent collection. Moreover, few patients treated with PORP could be included, whereas at other centers PORPs are more frequently used.

In summary, an average ABG of 20 dB (500–4000 Hz) should be considered in primary cholesteatoma cases with incus removal but intact and mobile stapes after one-stage ossicular reconstruction. Hearing after type III tympanoplasty was worse than after incus interposition or stapes augmentation. No difference in hearing between closed cavity incus interposition or stapes augmentation, and open cavity stapes augmentation was observed. The beneficial effect of ossicular reconstruction seemed limited to the lower and middle frequencies (up to 2 kHz), whereas the higher frequencies did not improve compared to the preoperative audiogram. Selecting the type of surgery between open and closed cavity techniques relies not primarily on the attempted hearing outcome but should have in mind the main surgical goal of complete disease removal. This is supported by our low recidivism rate of less than 10%.