Quantitative assessment of cochlear and vestibular ganglion neurons in temporal bones with chronic otitis media

Abstract

Purpose:

In this study, we aimed to determine whether or not COM leads to loss of spiral and Scarpa ganglion neurons.

Methods:

From the human temporal bone (HTB) collection at the University of Minnesota we selected human temporal bones with COM, defined as the presence of clinically intractable tissue abnormalities in the middle ear (cholesteatoma, perforation of the eardrum, granulation tissue, fibrosis, tympanosclerosis, and cholesterol granuloma). We also selected HTBs from donors with no ear diseases as controls. We quantitatively analyzed the number of spiral and Scarpa ganglion cells and compared the results obtained in the control and study groups.

Results:

In both COM and control groups, we observed a significant negative correlation between age and number of both spiral (R= −0.632; P<0.001; 95% CI= −0.766 – −0.434) and Scarpa ganglion (R=−0.404; P=0.008; 95% CI= −0.636 – −0.051) cells. We did not find any significant differences in the number of spiral ganglion cells (in total or per segment) or in the density of Scarpa ganglion cells (in each vestibular nerve or both) in the COM group as compared with controls (P>0.05).

Conclusions and relevance:

Our results did not demonstrate significant loss of cochlear or vestibular peripheral ganglion neuron loss in HTBs with COM as compared with controls.

Introduction

Otitis media is a major public health concern worldwide. [1, 2] The World Health Organization (WHO) estimated the global prevalence of chronic suppurative otitis media between 65–330 million individuals. [1] Otitis media associates with several long-term sequelae, including hearing loss, tinnitus, and vestibular problems, which impact negatively in the quality of life as it may lead to social, educational, and work-related problems. [1, 3–6] The hearing loss associated with otitis media may lead to communication problems, social isolation, academic underachievement, and unemployment. [7, 8] Furthermore, recent evidence has associated hearing loss at midlife as one of the most important modifiable risk factors for dementia. [9, 10]

Several authors studied the pathophysiologic mechanisms involved with inner ear sequelae secondary to otitis media. [7, 11, 12] In this regard, it has been observed that the round window membrane is permeable to inflammatory mediators, bacterial toxins and even whole bacteria, which may translocate from the middle ear to the inner ear. [13–15] Furthermore, studies in human temporal bones (HTBs) shown that chronic otitis media (COM) associates with significant loss of cochlear hair cells, atrophy of the stria vascularis, and loss of spiral ligament fibrocytes in the basal turn of the cochlea. Clinically, it was observed that the location of those inner ear abnormalities correlated tonotopically with the sensorineural hearing loss affecting the high frequencies in patients with COM. [8, 11, 12, 16] Recently, histopathologic studies have also shown significant loss of vestibular hair cells in HTBs with COM, suggesting that the vestibular neuroepithelium may be affected by COM as well. [16–19]

Although several studies demonstrated in the past the cochlear pathologic changes to the sensorial epithelium of the cochlea, research aimed to demonstrate abnormalities affecting the neural structures of the cochlea (spiral ganglion cells and nerve fibers) in temporal bones with chronic otitis media are scarce. [12, 20] Furthermore, to our knowledge, no previous study has yet dedicated to analyzing potential abnormalities affecting the vestibular ganglion neurons. Thus, the objective of our research is to quantitatively analyze the number of spiral and Scarpa ganglion neurons in HTBs with chronic otitis media and to compare the results to nondiseased HTBs.

Methods

Setting and sample selection

From the HTB collection at the University of Minnesota we selected temporal bones with COM. COM cases were defined based on characteristic histopathological findings, which is the presence of clinically intractable tissue abnormalities in the middle ear (such as cholesteatoma, perforation of the eardrum, granulation tissue, fibrosis, tympanosclerosis, and cholesterol granuloma) (Fig. 1). We excluded HTBs from donors who had (1) tumors affecting the head and neck or underwent chemo or radiation therapy of the head and neck; (2) hematologic neoplasms; (3) history of aminoglycoside use (either topical or systemic); (4) otologic surgery other than tympanostomy tubes; (5) Meniere’s disease; (6) clinical otosclerosis or histopathologic foci affecting the cochlear or vestibular endosteum; (7) perilymphatic fistula; or (8) systemic autoimmune disease. We also excluded temporal bones with severe removal or processing artifacts. Also, we selected HTBs from donors with no clinical history (in their clinical records) or histopathological signs of ear diseases. Those HTBs were age- and sex-matched (as closely as possible) to the HTBs in the COM group and included in a control group.

Fig. 1:

A horizontal human temporal bone section from a donor with chronic otitis media (Hematoxylin and eosin). (A) panoramic view at the epitympanic level, showing the presence of a tympanic membrane (TM) retraction, and the presence of fibrous and granulation tissue in the middle ear (squared area). In B, several tissue abnormalities are seen, including erosion of the short process of the incus, fibrosis, granulation tissue, and presence of serous-purulent effusion. EAC, external ear canal; TM, tympanic membrane; ME, middle ear; M, malleus; I, incus; FN, facial nerve; C, cochlea; V, vestibule; IAC, internal auditory canal.

All temporal bones we used had previously been harvested during autopsy, fixed in 10% buffered formalin, decalcified with ethylenediaminetetraacetic acid or trichloroacetic acid, dehydrated in graded concentrations of alcohol, and embedded in celloidin. Each bone was serially sectioned in the horizontal plane at a thickness of 20 mm, and every 10th section was stained by hematoxylin-eosin.

Two of the authors were responsible for scrutinizing the temporal bones and individually counted the Spiral and Scarpa ganglion cells. The authors were blinded from clinical and demographic information from the donors and performed the cell counting three times. The results obtained by both authors were then compared to ensure interobserver agreement.

The Institutional Review Board of our institutions (0206M26181 and 1.751.916) approved this study.

Middle ear abnormalities

In HTBs with COM, we analyzed the presence of abnormalities affecting the tympanic membrane (increased thickness, tympanosclerosis, retraction, perforation), ossicular chain (remodeling, erosion), middle ear mucosa (hyperplasia), and presence of fibrosis, granulation tissue, cholesteatoma or cholesterol granuloma, inflammatory cells. We also assessed the presence and type of middle ear effusion, which was classified in either serous or mucoid based in their histologic characteristics:[21] serous effusion was defined as the presence of a pale, yellow exudate with low viscosity; and mucoid effusion was characterized as the presence of a thick, opaque exudate.

Analysis of the neural structures of the cochlea

To analyze the number of spiral ganglion cells in the Rosenthal’s canal (Fig. 2), we used the methodology proposed by Otte et al. [22] The Rosenthal’s canal was divided in 4 segments – I (from base to 6 mm); II (6 to 15 mm); III (15 to 22 mm) and IV (22 mm to apex). To count the cells, we used a light microscope (Nikon Eclipse E100; Nikon Co., Tokyo, Japan) with a high-resolution camera attached (Nikon DS-Fi3; Nikon Co., Tokyo, Japan). The images were captured at a 400x magnification and then transferred to an image viewer software (NIS elements viewer – Nikon Co.; Tokyo, Japan). The computer software allowed labeling of each cell individually, reducing the risk of overestimation or underestimation of the number of cells. The total number of spiral ganglion cells in each segment was then multiplied by 10 (to account for cells in the transition between slides), and later corrected by a predefined factor of 0.9. [22]

Fig. 2:

A horizontal human temporal bone section from a donor with chronic otitis media (Hematoxylin and eosin). (A) panoramic view of the external, middle and inner ears. In the middle ear, mucoid purulent effusion with inflammatory cells is observed along with fibrosis and granulation tissue. (B) shows the cochlea in a higher magnification (4x). The squared area in (B) is seen in (D), showing the Rosenthal’s canal in the basal turn of the cochlea (segment II) (10x). (C) shows Scarpa ganglion neurons in the inferior vestibular nerve (10x).

Analysis of the neural structures of the vestibule

The density of Scarpa ganglion cells in the inferior and superior vestibular nerves (Fig. 2) was obtained using the methodology proposed by Richter. [23] Using a light microscope, in all 10th HTB section which included the superior and inferior vestibular nerves, we outlined the Scarpa ganglion cells with identifiable nucleoli, at a magnification of 400x. We captured the images using a camera attached to the light microscope, and then transferred to a computer software which allowed individual labeling of cells. All stained sections including the superior and inferior vestibular nerves were sequentially scrutinized, and we counted every Scarpa ganglion cell. The boundary between the superior and inferior nerves was determined based on the assumption that – considering that both nerves were cut perpendicularly – the number of ganglion cells increases (superior nerve), then decreases to a minimum (limit between superior and inferior nerve) before increasing again (inferior nerve). To avoid overestimation of data, we corrected the raw number of cells using the formula of Abercrombie [24] (correction factor = 0.88); then, the number of cells was also multiplied by 10 to account for unstained sections.

Statistical analysis

We used the Shapiro-Wilk test to assess our results regarding distribution, which revealed that all of our data was normally distributed. Therefore, to compare the number of cochlear and vestibular ganglion cells among groups, we used the parametric Student’s T-test. To analyze the influence of middle and inner ear abnormalities in the loss of cells in the COM group, we also used Student’s T-test and corrected for age differences using a general linear model and Bonferroni correction. All statistics were performed using SPSS 23.0 software for Windows (SPSS Inc., Chicago, IL). Findings were considered statistically significant when the P-value was less than 0.05.

Results

Demographics

The final COM group comprised of 31 HTBs from 29 donors. The mean age of the donors was 60.87 years (Median, 63; Range, 23–95; standard deviation [SD], 17.2); 21 (67.7%) specimens were from men and 10 (32.3%) from women. The control group also comprised of 31 HTBs from 24 donors. Their mean age was 60.32 (Median, 65; Range, 26–85; SD, 16.95), and 16 (51.7%) specimens were from men and 15 (48.3%) from women. There were no significant differences in the distribution of age (P=0.90; 95% confidence interval [CI] = −9.48 – +8.38) or sex (P=0.196) between our study and control groups.

We found audiograms available in the medical records of 6 (20.6%) donors from our COM group. All audiograms revealed the presence of mixed hearing loss. The degree of hearing loss ranged from mild to severe (mild, n=1, 16.6%; moderate, n=2, 33.3%; severe, n=3, 50.1%), and the mean pure-tone average (PTA) was 57.5 dB (median, 60.0 dB; range, 30 dB – 75 dB).

We found (in the medical records) of the donors with COM clinical information available for 22 (75.8%). In 18 (81.8%) of those medical records, we found descriptions of the presence of clinical COM. Among those 22 medical records containing clinical information, an objective description of the presence of hearing loss was present in 17 (77.2%), and the presence or absence of tinnitus was available in 13 (59.0%). Based on those medical records from which clinical information was available, the prevalence of hearing loss was estimated at 100% (n=17) and the prevalence of tinnitus at 76.9% (n=10). We also found (in the clinical records) information regarding the presence or absence of vestibular symptoms for ten donors (45.4%) – of those, 5 (50%) complained of vestibular symptoms. In the control group, we found information regarding hearing loss, tinnitus, and dizziness for ten donors (41.6%) (12 specimens) – of those, 3 (30%) complained of hearing loss, 2 (20%) had tinnitus, and 1 (10%) had episodic positional dizziness.

Middle ear findings

The pathologic changes and abnormalities we found in the HTBs with COM are in Table 1. The most frequent findings were the presence of tympanic membrane abnormalities (thickening, tympanosclerosis or perforation) (90.3%), presence of granulation tissue (90.3%), ossicular chain abnormalities (remodeling, erosion) (70.97%), and the presence of fibrosis (70.97%). In the majority (90.3%) of the HTBs with COM, we observed the presence of middle ear effusion, and the most frequent type was serous (61.9%). In the inner ear, we found saccular hydrops in 58.1% of the HTBs; in the cochlea, the most frequent abnormal finding was the collapse of Reissner’s membrane (n=6; 19.3%). Among HTBs from the control group, the middle ear was unremarkable in all 31 (100%) specimens; in the inner ear, we observed mild saccular hydrops in only one specimen (3.2%), and none (0.0%) had signs of cochlear hydrops.

Table 1.

Middle and inner ear abnormalities observed in temporal bones from the COM group

Abnormalities		N	%
Tympanic membrane	Yes	28	90.32%
	No	1	3.23%
	Artifacts	2	6.45%
Ossicular chain	Yes	22	70.97%
	No	7	22.58%
	Artifacts	2	6.45%
Hyperplastic mucosa	Slight	13	41.94%
	Moderate	8	25.81%
	Severe	5	16.13%
	No	5	16.13%
Fibrosis	Yes	22	70.97%
	No	9	29.03%
Granulation tissue	Slight	13	41.94%
	Moderate	8	25.81%
	Severe	7	22.58%
	No	3	9.68%
Cholesterol granuloma	Yes	3	9.68%
	No	28	90.32%
Cholesteatoma	Yes	5	16.13%
	No	26	83.87%
Effusion	Serous	19	61.29%
	Mucoid	9	29.03%
	No effusion	3	9.68%
Cochlear hydrops	Mild	1	3.23%
	Moderate	1	3.23%
	Severe	0	0.00%
	No	21	67.74%
	Collapse/artifact	7	22.58%
Saccular hydrops	Mild	12	38.71%
	Moderate	6	19.35%
	Severe	0	0.00%
	No	7	22.58%
	Collapse	1	3.23%
	Membranes broken/artifact	5	16.13%
Utricular hydrops	Mild	4	12.90%
	Moderate	0	0.00%
	Severe	0	0.00%
	No	26	83.87%
	Artifacts	1	3.23%
Labyrinthitis	Serous	15	48.39%
	Suppurative	0	0.00%
	Ossificans	0	0.00%
	No	16	51.61%

Quantitative assessment of Scarpa and Spiral ganglion cells

In both COM and control groups, we observed a significant negative correlation between age and number of both spiral (R= −0.632; P<0.001; 95% CI= −0.766 – −0.434) and Scarpa ganglion cells (R=−0.404; P=0.008; 95% CI= −0.636 – −0.051) (Fig. 3). We did not find significant correlations between the number of spiral or Scarpa ganglion cells and the presence of hearing loss, tinnitus, or vertigo, and degree of hearing loss or PTA (P>0.05). We did not observe a significant correlation between the number of cochlear or vestibular ganglion cells and the type of effusion, presence or absence of mucosal hyperplasia, fibrosis, granulation tissue, cholesteatoma, labyrinthitis, or hydrops. (P>0.05).

Fig. 3:

Scatterplot graphic with trendlines showing a decrease in the number of Spiral and Scarpa ganglion cells according to age in both chronic otitis media (COM) and control groups

We did not find any significant differences in the number of spiral ganglion cells (in total or per segment) or density of Scarpa ganglion cells (in each vestibular nerve or both) in the COM group as compared with controls (P>0.05) (Fig. 4).

Fig. 4.

Boxplot graphic representation of the number of spiral and Scarpa ganglion cells in the human temporal bones from the chronic otitis media (COM) and control groups

Discussion

There is abundant evidence in the literature demonstrating inner ear damage secondary to otitis media. [7, 8, 11, 12, 16, 17, 25] Although the exact effects of otitis media in the cochlear and vestibular structures are not completely understood, [12, 13, 26] many studies consistently demonstrated that COM associates significantly with cochlear hair cell loss, stria vascularis atrophy, and loss of fibrocytes of the spiral ligament.[16, 19, 26] Many potential pathophysiologic mechanisms have been proposed, including (1) passage of inflammatory mediators and bacterial toxins to the inner ear through the round window membrane, causing direct injury to the neuroepithelium; [13–15, 20, 20, 26, 27] (2) presence of an intracochlear response against inflammatory mediators through activation of pro-inflammatory genes and expression of cytokines; [28, 29] and (3) secondary endolymphatic hydrops. [11, 20, 30] More recently, it has been demonstrated that COM may also lead to significant hair cell loss in vestibular organs as well, more remarkably in the macula of the otolithic organs.[16–19] However, studies aiming to analyze potential abnormalities affecting the peripheral neural structures in HTBs with COM are scarce, and past studies in this regard yielded inconsistent results. [12, 19, 20]

We did not find a significant loss of spiral ganglion cells in any segment of the Rosenthal’s canal in HTBs with COM as compared with controls, which is consistent with the majority of studies in this regard. [12, 20] As previously observed, the only factor that significantly associated with a decrease in the number of spiral ganglion neurons was age. [22, 23, 31, 32] Although some authors demonstrated loss of spiral ganglion cells in HTBs with COM, those were mostly observed in HTBs from donors with COM who had either suppurative labyrinthitis or cochlear ossification, [19, 33] which were not present in any of the HTBs in our casuistic. Some HTBs in the COM group had signs of serous labyrinthitis, which also did not associate with spiral ganglion cell losses as compared with ones without serous labyrinthitis. [33] Therefore, our results suggest that the presence or absence of inflammatory cells, type of effusion, abnormalities in the ossicular chain, presence of cholesteatoma or cholesterol granuloma, presence of hydrops, and presence of serous labyrinthitis do not associate with significant loss of spiral ganglion cells in HTBs with COM. Additionally, our results suggest that the absence of significant loss of peripheral cochlear ganglion neuron secondary to COM (as compared with controls) would support maximum benefit for cochlear implantation in cases presenting with severe or profound hearing loss. [8]

To our knowledge, our study is the first to analyze the number of Scarpa ganglion cells in HTBs with COM. Recently, studies have shown that COM leads to significant loss of vestibular hair cells, especially in the otolithic organs; [16–19] therefore, we hypothesized that the vestibular ganglion neurons could be affected by COM as well. However, our results do not demonstrate a significant decrease in the number of Scarpa ganglion neurons in HTBs with COM as compared with controls. In both COM and control groups, we found a negative correlation between the number of vestibular ganglion cells and age of the donors, a trend that was also observed in previous studies. [23] Although we observed a higher prevalence of saccular hydrops in HTBs with COM as compared with controls, the presence of hydrops also did not associate with significant loss of Scarpa ganglion cells.

In conclusion, our findings do not support the assumption that COM may lead to loss of peripheral cochlear and vestibular ganglion cells; therefore, it seems that the significant loss of cochlear and vestibular hair cells previously observed in HTBs with COM do not associate with direct or retrograde ganglion degeneration. [34] However, in the cochlea, it has been previously observed that spiral ganglion neurons can survive for extended periods after significant loss of inner hair cells, [32, 35] while dendritic nerve fibers seem to be more prone to degeneration in cases of atrophy of the Organ of Corti. [34] Thus, our results may not necessarily denote a complete absence of peripheral neuron degeneration in HTBs with COM, as problems such as loss of dendritic fibers or abnormalities in the synapsis between distal fibers and the inner hair cell could not be assessed with our current methodology. In this regard, an experimental study has demonstrated that pure chronic conductive hearing loss – which is a hallmark of COM – may lead to remarkable abnormalities in the efferent and afferent innervation of cochlear hair cells. [36, 37] Considering that COM – in addition to the conductive deficits [7, 11, 12, 16] – also associates with loss of cochlear hair cells and sensorineural hearing loss, it is possible that COM may also lead to similar (or even worse) abnormalities in the distal cochlear nerve. Furthermore, some hearing problems frequently observed in patients with COM (such as tinnitus [3, 11, 38, 39] and sound discrimination problems [7, 40, 41]) have been also previously associated with neural dysfunction and cochlear synaptopathy. [42–44] The characterization of the progression and exact sites of cochlear and vestibular lesions secondary to COM is critical to allow the development of future interventions (such as stem cell therapy) designed to prevent or even regenerate those inner ear sequelae. [16, 17, 21, 33, 45] Therefore, we believe that further studies aiming to assess cochlear and vestibular dendritic nerve fibers and distal synapses may shed additional light on whether COM does lead to peripheral neural abnormalities in both cochlea and vestibule or not.

Our study has limitations. The lack of more extensive clinical information regarding hearing loss, tinnitus, and vestibular symptoms, as well as the scarcity of audiograms from the donors with COM, prevented a more in-depth statistical analysis aimed to correlate the number of ganglion cells with the presence of symptoms and audiometric data. Despite those limitations, our study provides the most substantial evidence demonstrating that COM does not associate with statistically significant losses of spiral and Scarpa ganglion neurons.

Conclusions

Our results did not demonstrate significant loss of cochlear and vestibular peripheral ganglion neurons in HTBs from donors with COM.