Abstract
Objectives:
The aims of our study were to investigate the clinical and audiometric outcome of the surgical treatment of postinflammatory medial meatal fibrosis (PIMMF) and to review the histopathologic changes in the specimens of the fibrotic plug, in order to try to shed light on the pathogenesis of the disease.
Materials and Methods:
The clinical records and the histopathologic specimens of all patients who underwent tympanomeatoplasty for PIMMF at the ENT Clinic of the University of Erlangen between 2006 and 2020 were evaluated retrospectively.
Results:
Thirty-four patients (41 primary surgical procedures) made up our study cohort. Of this, 28 cases were managed by means of meatoplasty and 13 cases with tympanomeatoplasty. The mean preoperative air–bone gap (ABG) was 27.8 dB (10-44 dB). Postoperative ABG was significantly improved compared to preoperative values at both short- and long-term follow-ups (P < .001 for both). No significant difference was noted between short-term and long-term ABG (P = .240). An ABG ≤20 dB was achieved in 65.8% of patients (short term) and 50% (long term). The overall rate of revision surgery for restenosis was 29.3% (12/41). Histopathologic reevaluation of the fibrotic plugs revealed a mosaic of patterns with frequent occurrence of secondary cholesteatoma-like lesions and keloid-like tissue changes. Lichenoid submucosal inflammation and increased ectopic ceruminous gland lobules were seen less frequently.
Discussion:
The moderate long-term outcome of surgical management and the identification of histologic changes with therapeutic implications might pave the way for alternative nonsurgical treatment options.
Introduction
The use of the term “idiopathic” in medicine relates to any disease or condition for which the cause is still unknown. The progressive change of the terminology from “idiopathic” to “postinflammatory” for medial meatal fibrosis is intended to describe the onset of the disease rather than to explain its pathogenesis. A considerable amount of literature has already dealt with this demanding entity, which is typically characterized by repetitive episodes of excessive granulation of the affected meatal skin of the bony ear canal with chronic secretion, fibrinoid degeneration, and reepithelialization with resulting fibrosis. 1 For the otologic surgeon, postinflammatory medial meatal surgery (PIMMF) suggests a demanding topic with a still unknown pathogenesis and frequently discouraging long-term surgical results.
The aim of our study was to investigate the audiometric and the surgical outcome (in terms of recurrence) in our cases with PIMMF and review the literature for relevant information on these clinical aspects. A further purpose of this study was to critically review the histopathologic changes in the specimens of the fibrotic plug in a trial to shed light on the pathogenesis of the disease.
Materials and Methods
This study was conducted at an academic tertiary referral center performing more than 700 otologic surgical procedures annually (…). The records of all patients treated for PIMMF between 2000 and 2020 were studied retrospectively. All patients were evaluated preoperatively by clinical examination, photographic documentation of the otoscopic finding, and by pure tone audiometry. Specifically, air and bone conduction thresholds at 0.5, 1, 2, and 4 kHz were used to calculate 4-tone pure-tone averages. The air–bone gap (ABG) was calculated as air conduction pure tone audiometry minus bone conduction pure tone audiometry. All histopathologic specimens were reexamined to detect potential signs of an underlying dermatologic disease. The tissue in our specimens was formalin-fixed and paraffin-embedded before cutting 1-µm sections and stained with hematoxylin and eosin. All specimens were assessed for the pattern of fibrosis (fine-reticular or dense collagenized keloid-like), grade of interstitial inflammation, dominant inflammatory cell type, lichenoid subepithelial inflammation, ectopic occurrence of apocrine (ceruminous) glands, content of elastic collagen fibers (using Verhoeff-Van Gieson elastic staining [EVG]), presence and quality of surface epithelium, as well as the secondary occurrence of cholesteatoma. The institutional review board of the University Hospital of Erlangen approved this study.
Statistical analysis was performed using the Wilcoxon signed-rank test with 95% CIs. This nonparametric test was chosen as the most suitable for analyzing repeated measurements on a single small-sized sample. The software SPSS version 21 for Windows (SPSS, Inc) was used for the analysis. A P value of <.05 was considered statistically significant.
Surgical Technique
After induction of general anesthesia and orotracheal intubation, local anesthetic (ultracain 2% with adrenalin 1:200 000) is infiltrated in a subperiostal plan in the postauricular region and the cartilaginous canal (“4-quadrant”). A split thickness skin graft (0.2 mm) is harvested from the postauricular region and placed in normal saline solution for use later in the procedure. An endaural incision is performed in the area between the tragus and the crus of helix, and a self-retaining retractor is placed. The soft tissue is elevated to the level of the bony ear canal and the skin is transsected just lateral to the granulation tissue and fibrous plug, causing the obliteration of the medial part of the ear canal. The granulation and fibrous plug are removed working from lateral to medial down to the level of the tympanic membrane, and scary soft tissue is dissected off the fibrous layer of the ear drum, which should ideally be left intact. If a perforation is created during this dissection or in the case of preexisting chronic otitis, tympanoplasty can be performed using a variety of standard techniques. A bony canaloplasty is routinely performed by means of a diamond burr to remove all canal wall bulges or overhangs as well as to broaden the ear canal. The split thickness skin graft is cut into 2 separate grafts. One graft is placed on the anterior wall and the other on the posterior wall of the ear canal. Medially, both grafts overlay the tympanic membrane. Silicone stripes stabilize the skin grafts in contact with the underlying bone, and Gelfoam is used to pack the ear canal. Finally, the endaural incision is closed.
Results
Thirty-four patients made up our study sample (17 men and 17 women). Forty-one primary surgeries were performed (27 patients unilaterally, 7 patients bilaterally). Their mean age at the time of first surgery was 55.2 years (range: 20-79 years). Postinflammatory medial meatal fibrosis was surgically managed by means of an endaural approach in 39 cases and a retroauricular approach in 2 cases. Twenty-eight cases were managed by means of meatoplasty, and the remaining 13 cases by tympanomeatoplasty.
The mean preoperative ABG was 27.8 (10-44 dB). Complete postoperative short-term audiometric data (mean: 6 months, range: 3-12 months) were available for all 41 ears, whereas long-term data (mean: 20 months, range: 12-160 months) were available for 22. The postoperative ABG was significantly improved compared to preoperative values at both the short-term (P = .001) and long-term follow-ups (P = .008). No significant difference was noted between the short-term and long-term ABG (15.5 ± 10.1 dB and 19.4 ± 10.8 dB, respectively, P = .240). Surgical success, defined as an ABG ≤20 dB, was achieved in the majority of patients (65.8% short term and 50% long term). Although all 41 ears had short-term clinical data, 22 had long-term clinical data available for review. The mean follow-up for the long-term cohort was approximately 18 months (12-48 months). Significantly, more patients had some degree of canal restenosis (partial to complete) at long-term follow-up (40.9%) when compared with short-term follow-up (7.3%, P = .000). The overall rate of revision surgery for restenosis was 29.3% (12/41). One patient underwent a second surgical revision for restenosis on the right side. Thirty histologic specimens were available for reevaluation. The results of the histopathologic reexamination are presented in Table 1. Representative examples of the main histopathologic findings are shown in Figure 1.
Histopathologic Changes in the Specimens of Our Study Sample.
Representative examples of histopathologic findings in postinflammatory medial meatal fibrosis (PIMMF) specimens (A, B, C, E, F: hematoxylin and eosin [H&E] stains; D: EvG stain). A, Secondary cholesteatoma; (B) prominent ceruminous gland lobules; (C) prominent angiofibromatous tissue with focal, predominantly lymphocytic, chronic inflammatory infiltrate; (D) prominent fibroelastosis (EvG); (E) prominent subepithelial chronic lichenoid inflammation, note luminal hyperkeratosis; (F) higher magnification of E showing sprinkling of lymphocytes into the basal epithelium.
Discussion
In contrast to acute external otitis, postinflammatory auditory canal stenosis is rare. As a synonym for this entity, terms such as “stenosing otitis externa,” “medial meatal fibrosis,” “obliterative otitis externa,” “postinflammatory acquired auditory atresia,” “postinflammatory medial meatal fibrosis,” or “acquired medial auditory canal fibrosis” have been used.2,3 All of these names have in common the inflammatory component (“otitis”) and the reference to a fibrotic nature of the stenosis. The exact pathomechanism of this entity is not yet clear. It remains controversial whether this inflammatory component derives from chronic suppurative otitis media (with perforation of the tympanic membrane) or chronic myringitis or whether it is simply a skin disease with involvement of the medial part of the external auditory canal and the cutaneous layer of the tympanic membrane (“stratum cutaneum”). Irrespective of the precise “initiating” etiology, the common end result of all possible causes of this enigmatic entity is a paucicellular granulating inflammation of the outer part of the eardrum and the medial part of the auditory canal with slowly progressive transition to fibrotic plaque and the subsequent development of fibrous obliteration. Involvement of the anterior tympanomeatal angle could lead to a severe conductive hearing loss. Furthermore, the progressive narrowing of the ear canal predisposes to further infections, transforming the whole process into a vicious circle. 3 Last but not least, the inclusion of epithelial elements in this scar tissue could lead to the formation of cholesteatoma with a considerable incidence (5.7%, 4 7.3%, 5 and 9% 6 ).
The main reasons for operating in PIMMF are the need to manage severe hearing impairment as well as to exclude secondary cholesteatoma. Since the introduction of the basic principles of surgical management of PIMMF in 19667,8 and the significant attempts to clarify its aetiology, 9 the philosophy and approach to treatment has barely changed. 2 Early surgery with complete removal of the fibrous plug, preservation of the lamina propria of the tympanic membrane, broadening of the medial part of the ear canal (removal of all canal wall bulges or overhangs), and coverage of the exposed bone with skin grafts of variable thickness should be generally included in every surgical concept. Interestingly, the only surgical modifications refer to the approach (transcanal, endaural, 2 and retroauricular 10 ) and to the thickness and origin of the skin used (full or split thickness skin grafts,11,12 preauricular or postauricular transpositional flaps.13,14) Undoubtedly, the most significant complication related to PIMMF is a restenosis of the ear canal. A meta-analytic investigation of this parameter is extremely complicated due to lacking information on the extent of stenosis (partial or complete restenosis with conductive hearing loss requiring revision surgery), the variable follow-up after primary surgery (“short-term” versus “long-term”) as well as the lacking or variable definition of the “long-term period” between different studies and working groups. Although the aforementioned surgical technique has been almost unanimously accepted and implemented, restenosis rates vary significantly between different studies. In general, although the short-term restenosis rate is still within relatively acceptable levels (5.7%, 6 12%, 15 and 22% 16 ), the overall impression is rather disappointing in the long term (7.9%, 6 13.8%, 10 25%, 16 33.3%, 17 and 60% 18 ). Our results did not prove to be an exception to this picture: Although 7.3% of our patients experienced an early restenosis and underwent immediate surgical revision, almost half of our study cohort had a clinically relevant recurrence of the stenosis in the long term (>12 months). For the sake of correctness, the interpretation of the results of our analysis should take specific factors into consideration: the retrospective design, the small sample size, and the modest proportion of the study patients available for long-term follow-up should be considered as limitations of the present study. These limitations could reduce the possibility of generalizing our outcomes with respect to the general population.
Over these years of witnessing rather discouraging long-term results, our clinical observations have led to several questions and self-critical considerations as part of the internal auditing system. Why does the fibrotic plug involve only the medial part of the external ear, practically never exceeding the bony-cartilaginous junction, and ending in the laterally still wide, medially blind-ended ear canal? On the other hand, why is this situation “frequently found especially among elderly patients who have been wearing a hearing aid for many years,” 9 that is, due to chronic irritation of more likely the lateral part of the ear canal? Furthermore, it seems that the description “postinflammatory” refers to the clinical onset of the disease but does not address the recurrent pattern as a feature of this disease. Does the cause of the disease depend on the quality of the skin of the medial part of the ear canal or the epithelial layer of the tympanic membrane or is the skin protected laterally from an extension of the inflammatory process by, for example, ceruminous glands? If the supposed underlying cause has been eliminated (eg, trauma or chronic otitis media), why does PIMMF recur? Is the etiology of the disease more likely to depend on the quality of the skin (or skin graft used in case of surgical management of the disease)? For these reasons, we performed a thorough review of the relevant literature to search for information on a possible association between PIMMF and well-known dermatologic conditions. Moser et al describe an ectopic occurrence of adnexal structures (apocrine glands) within the bony EAC 1 in PIMMF cases, whereas scarce case reports point to a potential association of PIMMF with lichen planus.19-21
In order to illuminate pathophysiologic processes in this entity, we reexamined the histopathologic specimens of all PIMMF cases in our department. Our findings led us to some interesting conclusions. First, the general impression of a heterogeneous mosaic of histologic pictures, with several forms of fibrosis, grades of interstitial inflammation, and the presence of ceruminous glands and elastosis, reflects the variable patterns of local tissue reaction to a still unknown triggering factor. Cases with severe angiomatous changes, prominent elastosis (as seen in sunburnt skin, but also in scar tissue), or the significant occurrence of ectopic ceruminous glands as well as sporadic lichenoid histopathologic changes form a fascinating mosaic of histologic changes, all being included under the general term “postinflammatory medial meatal fibrosis.” It could be that a group of different pathologic entities are hidden behind the clinical picture of an acquired stenosis of the medial part of the ear canal. In any case, the inability to identify distinct stages of inflammation and subsequent postinflammatory fibrosis in our specimens points to the fact that “idiopathic” (instead of postinflammatory) would probably better describe this entity. Secondly, the dominant pattern of inflammatory changes was that of a minor predominantly lymphocytic interstitial inflammation, corresponding to the chronic nature of this entity, irrespective of the degree or extent of fibrosis. Thirdly, the presence of cholesteatoma in almost half of our study cases (46.6%), in statistically significant association with the presence of hyperkeratotic epithelium on the surface of the fibrotic plug (P = .008), justifies the opinion that surgical intervention is mandatory in all PIMMF cases, irrespective of the severity of the clinical symptoms (hearing loss). Furthermore, the presence of lichenoid histopathologic changes in 2 of our specimens (in patients already having lichen planus) is in agreement with the scarce literature reports presuming a potential pathogenetic association between these entities or a participation of this systemic disease in the formation of scar tissue in the ear canal.19-21 As first described by Moser et al, 1 the presence of ceruminous (apocrine) glands was noted in 26.6% of our samples, with extensive presence in the majority of these specimens, frequently associated with periglandular fibroelastosis. It could be that the ectopic occurrence of adnexal structures within the skin of the bony ear canal (due to histomorphogenic aberrations or acquired due to posttraumatic or infectious causes) may either predispose to ear canal stenosis or can also be seen as a reparative measure against the chronic inflammation and the process of fibrosis. Notably, we did not observe any temporal heterogeneity or multiphasic inflammation suggestive of chronological evolution as seen in some (mostly immune-mediated) systemic fibroinflammatory diseases such as immunoglobulin G (IgG)4-related fibrosis. Moreover, no prominent plasma cells or storiform fibrosis was noted, thus ruling out the possibility of an IgG4-related disorder. Due to these findings and the overall paucity of plasma cells, we did not test our cases for IgG4 expression.
Probably, one of the most impressive histopathologic findings in our study was the presence of dense keloid-like fibrosis in almost three-fourth of our specimens. The scenario of imbalance between the phases of inflammation, proliferation, and tissue remodeling, which is an accepted reason for keloid formation, 22 could possibly offer an explanation for PIMMF. The well-known tendency for multiple recurrences in keloids might represent a potential explanation for the natural course and clinical behavior of PIMMF. The presence of hyperplastic or hyperkeratotic epithelium in a significant number of our cases, as commonly seen in keloid disease, further justifies considering similarities in the pathogenesis and dignity of both diseases. Lastly, the at least moderate presence of elastic fibers on EVG staining in almost half of our study specimens (43.3%) is one of the common changes in these diseases. The therapeutic implication of these observations is the reasonable possibility of implementing the large number of established (eg, corticosteroids, low-dose irradiation, imiquimod, and 5-fluorouracil) and emerging (eg, calcium blockers and angiotensin-converting enzyme inhibitors)23,24 nonsurgical therapeutic modalities suggested for keloids, for example, in recurrent PIMMF cases.
Footnotes
Authors’ Note
K.M. and V.T. contributed equally to this manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
