Three-dimensional quantification of fibrosis and ossification after cochlear implantation via virtual re-sectioning: Potential implications for residual hearing

Abstract

Hearing preservation may be achieved initially in the majority of patients after cochlear implantation, however, a significant proportion of these patients experience delayed hearing loss months or years later. A prior histological report in a case of delayed hearing loss suggested a potential cochlear mechanical origin of this hearing loss due to tissue fibrosis, and older case series highlight the frequent findings of post-implantation fibrosis and neoosteogenesis though without a focus on the impact on residual hearing. Here we present the largest series (N = 20) of 3-dimensionally reconstructed cochleae based on digitally scanned histologic sections from patients who were implanted during their lifetime. All patients were implanted with multichannel electrodes via a cochleostomy or an extended round window insertion. A quantified analysis of intracochlear tissue formation was carried out via virtual re-sectioning orthogonal to the cochlear spiral. Intracochlear tissue formation was present in every case. On average 33% (SD 14%) of the total cochlear volume was occupied by new tissue formation, consisting of 26% (SD 12%) fibrous and 7% (SD 6%) bony tissue. The round window was completely covered by fibro-osseous tissue in 85% of cases and was associated with an obstruction of the cochlear aqueduct in 100%. The basal part of the basilar membrane was at least partially abutted by the electrode or new tissue formation in every case, while the apical region, corresponding with a characteristic frequency of < 500 Hz, appeared normal in 89%. This quantitative analysis shows that after cochlear implantation via extended round window or cochleostomy, intracochlear fibrosis and neoossification are present in all cases at anatomical locations that could impact normal inner ear mechanics.

Figure 1:. Representative 2D histological sections of the varying degree of intracochlear fibrosis and ossification after CI.

A: Case 10: only limited fibrosis(+), close to the electrode array (*). B: Case 13: Extensive fibrosis (+) in both scala tympani and vestibuli. The scala media (M) is free from fibrosis. C: Case 14: Near total ossification of the scala tympani (++).

Figure 2:. Generating 3D reconstructions and resections perpendicular to the cochlear spiral.

Example case 5. All scale bars are 1.0 mm. A: Alignment of the histological sections with the ‘Align Slices’ tool in Amira. For visual purposes, only 3 of the 40–50 stained sections for each specimen are shown. B. Segmentation of the different cochlear structures. Normal cochlear fluid, the electrode, fibrous tissue, and bone formation are segmented in blue, yellow, grey, and white, respectively. C: Based on the segmentations (panel B) a 3D reconstruction of the cochlear volumes is generated. At every 200 μm along the cochlear duct, a virtual box (red) is positioned at an angle perpendicular to the spiral. With the ‘Arithmetic Tool’, the overlap between this virtual box and the 3D cochlear reconstruction can be made. The result of this overlap is visualized in panel D. D: Virtual resection perpendicular to the spiral at the location visualized in C. The same color code as in B applies. For visual orientation, the original histological section is visible behind the virtual resection. SV/SM: Scala Vestibuli/Scala media; ST: Scala Tympani; E: Electrode; CP: Cochlear Partition

Figure 3:. Absolute volumes of new tissue formation in the scala tympani.

Each panel shows the results of the virtual re-sectioning of the 3D reconstructed cochleae. The cochlear volumes of these virtual sections are determined every 200 μm along the cochlear duct and indicated on the y-axis. The x-axis shows the cochlear frequency map as defined by Greenwood (1990). Normal cochlear fluid, fibrous tissue, and bone formation are plotted in blue, grey, and white, respectively. The electrode array location (the spiral extent, not the volume) is plotted in yellow. Because of the lack of a peri-electrode fibrous sheath in parts along the cochlear duct in cases 2, 5, and 9, the electrode volume is added to the normal perilymph volume at these locations, resulting in a potential overestimation of the perilymph volume without affecting the fibrosis or bone volumes. See supplemental figure 1 for plots with the angle from the round window at the X-axis.

Figure 4:. Tissue Formation in the Round Window Niche.

Upper Panel: For each case, the proportion of the round window membrane covered by either bone, fibrosis, or normal perilymph is plotted. Lower Panel: For the same corresponding cases, the occlusion of the cochlear aqueduct is visualized through the same color code with white and grey indicating bony or fibrous obstruction, respectively.

Figure 5:. Cochlear Aqueduct Occlusion Rate- CI cases vs Non-implanted ears.

Bars indicate the percentage of the sample where either a patent or obstructed cochlear aqueduct was histologically observed, as defined by Gopen et al 1997. (Gopen et al., 1997)

Figure 6:. Fibrosis and neo-ossification abutting the basilar membrane.

Each panel shows cochlear frequency on the x axis. For each point of the frequency map, at a resolution of 200 µm, the proportion of the basilar membrane that is lined by either bone, fibrosis, the electrode array, or normal perilymph is plotted. Note that at the extreme base, there are missing data (black shading), due to difficulties reconstructing the basilar membrane in the hook region. See supplemental figures 2-6 for plots with of the other parts of the cochlear partition (i.e. bridge, osseous spiral lamina) or with the angle from the round window at the X-axis.