Middle ear cholesteatoma is a mass formed by keratinizing squamous epithelium in the tympanic cavity and/or mastoid and subepithelial connective tissue and by progressive accumulation of keratin debris. Recurrent infections and inflammatory reactions within the subepithelial connective tissue by cholesteatoma contribute to bone resorption in the adjacent areas (1). Middle ear ossicles can thus be eroded, particularly the incus, then the stapes (2). Treatment of cholesteatoma is surgical. The primary goal of surgery is complete resection of cholesteatoma and avoidance of residual disease. The secondary goals are reconstruction (tympanic membrane, canal wall, atticotomy) and hearing restoration; it is now well established that ossiculoplasty can be performed as part of a single-stage operation (3, 4). In this regard, there are some knowledge gaps and discrepancies in the literature concerning hearing outcomes and their predictive factors in cholesteatoma surgery. This study aimed to analyze auditory results following cholesteatoma surgery with ossiculoplasty and identify potential prognostic factors related to middle ear status or surgical technique.

This is a retrospective study conducted on patients undergoing ossiculoplasty performed as part of cholesteatoma surgery between 2019 and 2024 (6 years) at our university medical center. The study population involved pediatric and adult patients requiring surgery for cholesteatoma associated with a non-continuous ossicular chain. Patients having cholesteatoma with a continuous ossicular chain and not undergoing ossiculoplasty were not included. Data were collected from paper medical records and computerized data, enabling the collection of the required clinical and audiometric information. Preoperative, as well as early (2 months) and midterm (6–9 months, median 8 months), postoperative audiograms were selected for analysis. We collected the puretone average (PTA) of every patient by calculating the mean air conduction (AC) and mean bone conduction (BC) at 500, 1,000, 2,000, and 4,000 Hz, thus adhering to the guidelines of the Committee on Hearing and Equilibrium for the evaluation of results of treatment of conductive hearing loss (5). We then evaluated the conductive hearing loss by calculating the mean air–bone gap: ABG = mean AC − mean BC. We defined the short-term and midterm auditory change (Δ) as the mean preoperative thresholds minus the mean short-term and midterm postoperative thresholds, respectively (ΔBC, ΔAC, and ΔABG). Patients lost to follow-up (not having at least one audiometry at 2 months or at 8 months) were excluded from the analysis. All patients were operated on in the same institution by the same experienced surgeon. The choice of ossiculoplasty depended on anatomical criteria, namely, the presence or absence of the stapes superstructure. We analyzed the functional results of ossiculoplasty in cholesteatoma and tried to identify pre- and postoperative parameters associated with audiometric data: age, surgical technique, presence or absence of the malleus handle and stapes superstructure, type of ossiculoplasty, occurrence of complications, mastoid obliteration, status of the middle ear mucosa, and primary or iterative nature of the procedure.

All variables were described using means (standard deviation, SD) and medians (quartiles Q1–Q3) for quantitative variables and sample size (N, %) for qualitative variables. The paired Student’s t-test was used to analyze the gain in air–bone gap depending on the frequency and for the comparison of short-term and midterm auditory results. The Kruskal–Wallis test was used to evaluate the significance of predictive factors of hearing outcome (non-normality of quantitative variables). The Student’s t-test was also used to assess whether midterm hearing results (BC, AC, and ABG) differed significantly from preoperative results. The risk α was set at 5%. The R software (version 4.2.1) was used for statistical analyses.

The primary goal of surgery for cholesteatoma is disease eradication, with hearing restoration as a secondary objective. The analyzed cases in this study involved patients who initially presented with an aggressive cholesteatoma with 28 complications before surgery. In fact, the present authors practice in an academic, tertiary referral center, and the referred patients often present with complex and challenging cases. This may explain the modest hearing gains observed in our study, especially considering that over one-third of our patients had undergone prior surgery, mostly in other institutions. This being said, our auditory results are comparable to those reported by other highly experienced surgical teams who reported on results of ossiculoplasties in cholesteatoma surgery (3, 4, 6), although they are slightly inferior to those reported in others (7–9).

Although not the primary focus of our study, it is worth mentioning that the overall recidivism rate in our series (16/88  =  18.2% at 1–3 years, all residual cholesteatomas) falls within the range reported in other studies—including both pediatric and adult patients—which varies from 1.4% to 23.4% over follow-ups of 13 to 36 months (5, 7, 9, 10).

If we consider only the adult population in our study, the residual rate was 14.7%; this is also comparable with the rate of recidivism (residual and recurrent cholesteatoma) reported in a study of adult patients only, which ranged from 12.7% to 21.3% over nearly 3 years of follow-up (10).

Now if we consider only the pediatric population in our study, the residual rate almost doubled (30%), which is a little less elevated than the rates of a study analyzing only pediatric patients, but with a much longer follow-up period (39% residual disease at 3 years and 45% at 6 years) (11).

It is also important to mention that the extrusion rate (1.1%) of titanium prostheses in our work was as low as other published series (0%–3.5%) using, among others, titanium prostheses, with longer follow-up periods (1–5 years) (6, 7, 9, 12–14).

Most studies focusing on auditory results after ossiculoplasty in cholesteatoma surgery do not mention the rate of surgical complications. Our rate is relatively low, given the high rate of preoperative complications (15).

We sought prognostic factors for hearing outcomes. The presence of the stapes superstructure had a positive influence on preoperative ABG. One interpretation is that when the stapes is still present preoperatively, the disease is less erosive (less osteolytic). Regarding the midterm postoperative gain in ABG or midterm “ΔABG,” the only factor positively influencing this parameter (bigger Δ) was the absence of posterior tympanotomy. This could be hypothetically explained by the fact that cases necessitating a posterior tympanotomy reflect more extensive disease in the mesotympanum and retrotympanum (sinus tympani). When considering the midterm ABG alone, the absence of mastoidectomy was the only predictive factor of good hearing results. This positive effect is probably related to less extensive disease, particularly in the epitympanum and antrum, which obviates the need for mastoidectomy. This was also found by other authors (16).

Considering the midterm “ΔAC,” the only predictive factor for a good hearing result (bigger Δ) was the primary surgery vs. revision surgery. This positive effect could be explained by the fact that revision surgery reflects recurrent and aggressive disease and yields more fibrosis and tympanosclerosis than primary surgery. This favorable effect of primary surgery was observed in other studies (7, 13, 16). When considering the midterm postoperative AC alone, there were three favorable predictive factors (for a smaller postoperative AC), namely, younger age, primary surgery, and absence of mastoidectomy. The positive effect of the younger age could be interpreted by the fact that younger patients already had better hearing (smaller AC) preoperatively.

It is worth noting that in our study, the presence or absence of the stapes superstructure had no significant effect on postoperative hearing results, similar to other studies (16, 17) and contrary to others (13, 18). Moreover, the presence or absence of the malleus handle did not have any impact on hearing results in our data, although the necessity of its removal was generally associated with more aggressive disease. This is also shared by some authors (6, 9) but contradicted by others (12, 14, 16, 19).

In the literature, predictive factors having an impact on hearing outcome after cholesteatoma surgery with ossiculoplasty are highly variable (3, 6–9, 20). Fukuda et al. found that in attic cholesteatoma, postoperative hearing outcome worsened with an increase in staging of the disease (I to II to III), according to the European Academy of Otology and Neuro-Otology/Japanese Otological Society (EAONO/JOS) classification and with the involvement of the stapes with cholesteatoma or granulation tissue (20). Using costal cartilage homografts, Quaranta et al. (8) noticed no difference between PORP and TORP groups in terms of postoperative ABG levels. Contrarily, using autograft and hydroxyapatite ossicular prostheses, Sevik Elicora et al. (3) found that PORPs yielded significantly lower postoperative ABG levels than TORPs did, but no difference was found with regard to hearing gain, probably due to the relatively better preoperative hearing levels in patients with an intact stapes.

Espitalier et al. (7) analyzed patients undergoing cholesteatoma surgery with an intact stapes and observed that the success rate (ABG ≤ 20 dB) was significantly higher in primary surgery than in revision surgery. Acke et al. (6) analyzed hearing results of ossiculoplasties in primary cholesteatoma surgery with an intact stapes; they found that incus transposition and CWU mastoidectomy resulted in a lower residual ABG after surgery. Querat et al. (9) analyzed hearing data of ossiculoplasties in CWU tympanoplasties for cholesteatoma with an intact stapes. They found no statistical difference in postoperative hearing results between the cartilage group (stapes augmentation) and the hydroxyapatite PORP group.

Some studies analyzed hearing results after ossiculoplasty and included other pathologies in addition to cholesteatoma, the latter being the most frequently encountered pathology, though. De Vos et al. (12) and Truy et al. (14) observed that the presence/absence of the malleus handle and the status of the middle ear mucosa were the only parameters that had a statistically predictive value on postoperative hearing. Furthermore, using the Austin–Kartush classification of ossicular status, De Vos et al. (12) noted that the best hearing results were obtained in Class B (malleus present M+, stapes absent S−), while Austin et al. (18) found a positive role if the stapes was present, in addition to the malleus. Schmerber et al. (13) noticed that the good prognostic factors for postoperative hearing restoration were the presence of the stapes superstructure, preservation of the canal wall during mastoidectomy (CWU), primary surgery, and the presence of the malleus handle. Gelfand and Chang (21) noticed that the main bad prognostic factor in all ossiculoplasties using titanium prostheses was cholesteatoma (PORP and TORP), as opposed to other middle ear pathologies; this was not found in hydroxyapatite prostheses. Dornhoffer and Gardner (16), in their analysis of ossiculoplasty in different pathologies, found that the absence of ear drainage, normal mucosa, presence of the malleus, absence of mastoidectomy, and primary surgery were statistically significant predictors of good hearing outcome. In a nice systematic review of the literature, Blom et al. (17) noticed that the only parameter predictive of hearing outcome in surgery for chronic otitis media with or without cholesteatoma was the malleus status: positive predictor if present, regardless of the stapes condition. Of note, they found no study analyzing the presence or absence of the incus.

Our work has many limitations. First, this is a retrospective study, as were the aforementioned studies. However, conducting a prospective—and particularly a randomized—study in patients with cholesteatoma is extremely challenging due to the wide variability in ossicular damage, disease extension, and staging. Second, when the stapes was present, different types of ossicular reconstruction were used, in contrast to other studies where a single type of prosthesis was consistently employed (7, 8, 13, 16, 20). We used chiefly stapes–cartilage assembly and PORP. This adds further heterogeneity in interpreting the results and stems from the retrospective design of the study and, more importantly, from the individualized approach to ossiculoplasty. Third, 76 out of 88 cases were analyzed at 6–9 months (median 8) postoperatively because 12 cases lacked audiometry at 8 months of follow-up, due to patients living far from our institution, being followed by another otologist, or returning only 12–18 months after surgery (for postoperative MRI). This is also due to the retrospective nature of the present study and thus the absence of a pre-established follow-up schedule. Fourth, we did not perform multivariate analyses but only univariate ones, because the main goal of our analyses was descriptive. In addition, the small number of significant variables (parameters) in univariate analysis as well as the small number of patients per variable and per modality was not in favor of multivariate analysis. Fifth and last, many variables (ΔABG, ABG, ΔBC, ΔAC, AC) were tested because these are the parameters that the otologists measure pre- and postoperatively. The high number of variables in a relatively small sample size can increase the risk of Type I error (risk of false positives) which is also a limitation of our study.

The auditory results after surgery for cholesteatoma were fair in our hands, at least in part because of the initial preoperative aggressiveness and extension of the disease. The rates of residual cholesteatoma and prosthesis extrusion were comparable to other studies published in the literature, and the rate of surgical complications was relatively low.

To summarize, a younger age and the presence of stapes superstructure were predictive factors of better preoperative hearing. Postoperatively, a younger age, the absence of mastoidectomy, the absence of posterior tympanotomy, and primary surgery were generally predictors of good hearing outcome. We found no effect from other factors, particularly the presence or absence of the malleus, which is often emphasized in other articles. Indeed, results in the literature are very diverse and sometimes contradictory. This stems from the highly variable degrees of disease extension and osteolysis, the surgical techniques including the ossiculoplasty, and the materials of ossiculoplasty used across studies.