Scanning electron micrograph showing the widespread distribution of micropores in the modiolar osseous spiral lamina (arrow) and floor (arrowhead) of the human scala tympani. The inset illustrates the anatomical orientation of this micrograph. There are many small micropores and a smaller number of larger openings in the osseous spiral lamina near the floor of the scala tympani. Note the relatively large number of openings in the modiolar side of Rosenthal's canal (RC) (double arrowhead), through which bundles of spiral ganglion neuron's central processes project to the internal auditory meatus. Bar = 100 µm.
Higher power scanning electron micrograph illustrating the fine structure of the canaliculae perforantes in the OSL adjacent to Rosenthal's canal in the basal turn of the human scala tympani. Note the large variation in canaliculae diameters. Bar = 10 µm.
Scanning electron micrograph illustrating the porous nature of the bone adjacent to the spiral ligament (see inset [arrow]) of the human scala tympani. Bar = 10 µm.
Scanning electron micrograph illustrating the medial surface of the scala vestibuli (see inset [arrow]) in the human cochlea. Unlike most of the perilymphatic surface in the scala tympani, there is minimal evidence of micropores in the scala vestibuli. Bar = 10 µm.
Mean diameter of canaliculi from the modiolar osseous spiral lamina in the human and the cat. The canaliculi diameter in the human was at least twice that in the cat. Error bar = 1 SD. These data are based on the following sample sizes for turns 1, 2, and 3 (human: n = 58, 71, and 30, respectively; cat: n = 54, 39, and 59, respectively).
Mean thickness of the modiolar osseous spiral lamina (OSL) in the human and the cat as a function of cochlear location. The human OSL is consistently thinner than the cat in all 3 cochlear turns. Both species exhibit a consistent reduction in OSL thickness apicalward. Error bar = 1 SD. These data are based on the following sample sizes for turns 1, 2, and 3 (human: n = 20, 19, and 23, respectively; cat: n = 4, 4, and 12, respectively).
Scanning electron micrograph illustrating the fine structure of the canaliculi perforantes in the osseous spiral lamina adjacent to Rosenthal's canal in the basal turn of the cat scala tympani (inset [arrow]). Note the reduction in canaliculi diameters and density compared with the human material. Bar = 10 µm.
Scanning electron micrograph illustrating the thickness of the osseous spiral lamina (arrowhead) adjacent to Rosenthal's canal (arrow) in the cat scala tympani. This bony wall, separating the scala tympani from Rosenthal's canal, not only contains numerous canaliculi but is also less than 40-µm thick in the cat. Note the evidence of micropores on the floor of the scala tympani (double arrowhead). Bar = 10 µm.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Shepherd RK, Colreavy MP. Surface Microstructure of the Perilymphatic Space: Implications for Cochlear Implants and Cell- or Drug-Based Therapies. Arch Otolaryngol Head Neck Surg. 2004;130(5):518–523. doi:10.1001/archotol.130.5.518
To study the surface microstructure of the scala tympani and scala vestibuli in humans and cats using scanning electron microscopy.
Cochleas from 8 humans and 4 cats were harvested and the otic capsule and soft tissue removed before the cochleas were prepared for scanning electron microscopy. Micrographs were taken of the bony surface of both the scala tympani and scala vestibuli in each cochlear turn. The diameter and density of the micropores (canaliculi perforantes) and the thickness of the osseous spiral lamina (OSL) adjacent to Rosenthal's canal was measured.
The human cochlea exhibits numerous canaliculi on the surface of the scala tympani, particularly associated with the OSL. There was a large range of diameters in the modiolar region of the OSL (0.2-23.0 µm). The OSL was also very thin, with a mean thickness of 26.8 µm in the base, tapering to 8.4 µm in the apical turn. Far fewer canaliculi were evident in the scala vestibuli. Examination of the cat cochleas showed a similar distribution of canaliculi to that seen in the human; however, they were smaller in diameter and the OSL was thicker than in the human cochleas.
The OSL is a thin and highly porous bony lamina that would appear to provide an open and extensive fluid communication channel between the scala tympani and Rosenthal's canal. These findings have important implications for the design and application of perimodiolar cochlear implant electrode arrays and may provide a potential route for drug- and cell-based cochlear therapies delivered via the scala tympani.
The microstructure of the osseous surface of the perilymphatic space within the mammalian cochlea has previously been studied to examine intercellular fluid pathways, particularly within the organ of Corti.1-5 Schuknecht and Seifi1 were the first to describe small pores on the scala tympani surface of the osseous spiral lamina (OSL) in cats. This work demonstrated that these bony pores, termed canaliculi perforantes, were commonly observed toward the peripheral region of the OSL close to the habenula perforata. Importantly, these studies demonstrated the presence of perilymphatic fluid channels within the perineural and perivascular spaces within the OSL that were considered to be the basis of communication of perilymph between the scala tympani and the tunnel of Corti.
The presence of these bony pores in the peripheral region of the OSL in cats were subsequently demonstrated in other species, including guinea pigs,2-4,6 chinchillas,5 and squirrel monkeys.5 A common feature of these studies was the large variability in pore diameter evident within a given species. Several authors have demonstrated the patency of canaliculi perforantes and the underlying perineural and perivascular spaces within the OSL in either live animals or fresh cadavers (ie, where there has been no change to the bone lining cells of the OSL).1,3,5,7
The development of a new generation of electrode array for cochlear implants based on their proximity to the modiolus8-10 and the potential clinical application of drug- or cell-based therapies for the protection or regeneration of hair cells and auditory neurons11-15 motivated our interest in examining the microstructure of the osseous surface of the perilymphatic space. We placed particular emphasis on examining the OSL in the region of the modiolus. In addition, we extended the previous work performed using experimental animals to include human material.
Eight formalin-fixed human temporal bones from cadavers of unknown sex and age were used in this study. The lateral aspect of each temporal bone was removed using a sagittal saw. The tympanic membrane was visualized and removed under magnification. The promontory and round window were then used as surface landmarks for the cochlea. The ossicles were removed, taking care to preserve the stapes footplate. The bone overlying the lateral aspect of the cochlea was gradually removed under magnification using initially cutting and then diamond burrs. Care was taken not to breach the perilymphatic spaces of the cochlea until all the bone had been "blue lined" to the level of the endosteum. This was to prevent bone dust from entering the cochlea. The lateral wall of the cochlea was finally removed under gentle irrigation. Throughout this procedure, care was taken to ensure that the underlying anatomic structures were left undisturbed.
As a comparative measure, 4 normal cat cochleas were also harvested. The cats were anesthetized with an intramuscular dose of ketamine (20 mg/kg) and xylazine (4 mg/kg), followed by a lethal dose of sodium pentobarbitol (100 mg/kg). The animals were systemically perfused with isotonic sodium chloride solution followed by 4% paraformaldehyde in 0.1M phosphate buffer (pH 7.4). Cat cochleas were harvested using the same microsurgical techniques as described for the human cochleas.
Soft tissue lining the perilymphatic surface was removed to fully evaluate the underlying bony structure. A similar approach was used in previous studies of this kind.5,16 Papain, an enzymatic digester, was chosen because it only digests the surface protein and not the underlying osseous structures. The fixed cochleas were washed in distilled water (4 times for 5 minutes) and placed in 10 mL of 10% papain for 2 weeks. The papain was changed at days 5 and 10. To control for the effects of papain, some cochlear samples were not subjected to this process. On completion of the protein digestion, each sample was rinsed in distilled water (4 times for 5 minutes) and dehydrated in serial ethanol concentrations of 30%, 50%, 70%, 80%, 95%, and 100% for 5 minutes each. The cochleas were then stored in 100% ethanol until processing for scanning electron microscopy.
Each specimen was placed in an oven and air dried at 60°C for 10 minutes. The specimens were then placed on an adhesive mount with a calibrated measuring grid and gold coated in an argon vacuum. Scanning electron microscopy was performed using a Phillips 515 (Phillips, Eindhoven, the Netherlands) scanning electron microscope using a magnification range of ×20 to ×50 000.
The proximal, basal, middle and apical turns of the human and cat cochleas were systematically photographed along the cochlear spiral. Both scala vestibuli and scala tympani were examined, with particular attention given to the OSL in the region of the modiolus adjacent to Rosenthal's canal. This location was selected because of its proximity to the optimal site of the perimodiolar electrode array. Calibrated photomicrographs of this region were used to measure the maximum diameter of each pore and their density (number of pores per square micrometer of modiolar OSL surface). The diameter of all clearly demarcated pores was measured and the area of the OSL determined using dividers and a ruler. Both pore diameter and density measurements were restricted to the central portion of each photomicrograph to ensure that only complete pores were included in the analysis. In samples where the OSL had fractured in the region of the modiolus, it was possible to measure the thickness of the OSL overlying Rosenthal's canal. Measurements were taken in all 3 cochlear turns and expressed as mean ± SD.
This study was conducted under the approval of the Royal Victorian Eye and Ear Hospital's Human Research Ethics Committee and Animal Research Ethics Committee.
Numerous micropores or "canaliculi" were evident throughout the length of the scala tympani (Figure 1). Most were associated with the OSL, both in the region of the modiolus and in its periphery close to the basilar membrane (Figure 2). Evidence of pores were also seen on the outer wall of the scala tympani close to the spiral ligament and, to a lesser degree, on the floor of the scala tympani (Figure 3). There was a marked difference between the surface ultrastructure of the scala tympani exhibiting numerous canaliculi and the osseous wall of the scala vestibuli that seemed virtually devoid of these structures (Figure 4).
There was a large range in the diameter of canaliculi in the modiolar region of the OSL (range, 0.2-23.0 µm; 5.2 ± 4.9 µm for the basal, 3.4 ± 3.0 µm for the middle, and 7.4 ± 5.2 µm for the apical turns; Figure 5). Of particular note, the mean diameter of canaliculi in the basal turn of the control specimens (ie, not subjected to papain) was 3.3 ± 3.1 µm and showed no statistically significant difference to that of the papain-treated specimens from the same cochlear turn (P = .15, t test). The density of the canaliculi (number of canaliculi per square micrometer of OSL area) remained relatively constant as a function of cochlear length. The density was 1.05 ± 1.03/µm2 in the basal, 1.43 ± 0.56/µm2 in the middle, and 1.25 ± 0.35/µm2 in the apical turns. The modiolar section of the OSL overlying Rosenthal's canal was revealed to be a very thin structure; the OSL thickness was 26.8 ± 6.0 µm in the basal, 14.5 ± 2.9 µm in the middle, and 8.4 ± 4.0 µm in the apical turns (Figure 6).
These porelike structures were widely distributed within the bony surface of the scala tympani and were also evident on the floor and outer wall of the human cochlea (Figure 1). The density of the pores at these sites was less than that at the OSL. The diameter of canaliculi in the outer wall and floor of the scala tympani were not measured in this study.
The distribution of canaliculi in the cat cochleas was similar to that observed in the human material, although the canaliculi within the OSL of the cat cochleas were of a smaller diameter (2.2 ± 1.6 µm in the basal, 1.3 ± 0.9 µm in the middle, and 1.5 ± 1.2 µm in the apical turns; Figure 5 and Figure 7) and less densely distributed (0.52 ± 0.01/µm2 in the basal, 1.2 ± 0.9/µm2 in the middle, and 0.56 ± 0.18/µm2 in the apical turns). The OSL overlying Rosenthal's canal was also found to be considerably thicker in the cat cochleas compared with the human material (OSL thickness, 40.0 ± 1.7 µm in the basal, 24.8 ± 1.8 µm in the middle, and 10.6 ± 1.7 µm in the apical turns; Figure 6 and Figure 8).
The present study has confirmed and extended the work of previous research examining the surface microstructure of the perilymphatic space. Importantly, this work has shown that canaliculi perforantes are more widely distributed than many previous reports had suggested; not only are they found in the peripheral OSL—the so-called zona perforata5—but extend to include the modiolar portion of the OSL. In addition to the OSL, evidence of micropores was also found on the floor and outer wall of the scala tympani. In contrast, we observed no clear evidence of canaliculi on the osseous surface of the scala vestibuli.
While demonstrating an extensive porous network in the bone lining the scala tympani, the present study has also highlighted the delicate nature of the OSL in the region of the modiolus. This was particularly true for the human cochlea, where the OSL thickness decreased from a mean of 27 µm in the basal turn to 8 mm in the apical turn. This finding highlights the need to ensure that cochlear implant electrode arrays and their insertion techniques are carefully designed to minimize contact with the OSL. Damage to this structure has been shown to result in extensive loss of spiral ganglion neurons (SGNs) in the region of the trauma.17
The apicalward reduction in the thickness of the OSL has also been reported in other studies. For example, Tinling and Chole18 have presented evidence of an OSL in the basal and middle turns of both rat and gerbil cochleas; however, they reported no OSL in the apical turn of these species. It would appear that apical SGNs in these species are directly exposed to perilymph.
A number of the present findings in human cochleas are consistent with a previous report by Küçük and colleagues16 describing the surface features of the human OSL. These authors also described a lacelike appearance of the OSL in the region of the modiolus, with pores of up to 45 µm in diameter, as well as fenestrations in the bony floor of the scala tympani. In addition, Küçük and colleagues16 described fenestrations of larger diameter in the floor of the scala tympani, which they suggested were associated with capillaries and collecting venules of the cochlear circulatory system. We observed similar structures (eg, Figure 1) and concur with the hypothesis by Küçük and colleagues16 that these structures are associated with the circulatory system.
Our feline data also revealed an extensive network of canaliculi consistent with earlier work in this species1; although it is apparent that the cat—and all other lower mammals examined to date—exhibit a smaller, less dense network of canaliculi than that observed in the human cochlea. Furthermore, the present study described an extensive canalicular network in the OSL close to Rosenthal's canal. While most previous animal studies have only reported canaliculi in the periphery of the OSL, 2 reports describe this network extending to the modiolar section of the OSL.2,3
Although the bony samples examined in the present study were typically devoid of soft tissue due to papain treatment, the OSL is normally in close contact with quiescent osteoblasts, termed bone lining cells.19 Bone lining cells have been identified in all species studied to date, including mouse, rat, gerbil, and human, and while they sometimes show processes that extend into the canalicular system,19 they are not extensive enough to prevent fluid communication channels between the scala tympani and the underlying perineural space. Ultrastructural studies have demonstrated that these cells appear to be closely applied to, but do not completely cover, the underlying bone matrix.4,18-20 For example, Tinling and Chole18 estimated that only approximately 12% of the modiolar OSL was covered by bone lining cells in the 3 mammalian species they studied.
There is good evidence from studies using several species that the canaliculi perforantes are functionally patent and provide an effective fluid communication channel between the perilymph of the scala tympani and (1) perilymph within the organ of Corti1,5; (2) the perineural space of the peripheral processes of the OSL and SGNs within Rosenthal's canal3-6; and (3) the cerebrospinal fluid.3,4,18 Furthermore, ultrastructural studies of the canaliculi perforantes using techniques in which bone lining cells over the OSL are left intact have concluded that the canaliculi perforantes generally provide direct fluid communication between the scala tympani and the perivascular and perineural spaces adjacent to auditory neurons.4,21 Several authors consider that the canaliculi play a role in supplying nutrients to the SGNs,3,5 while others have suggested that the network of pores located on the lateral wall of the scala tympani may play a role in fluid communication between perilymph and the spiral ligament.2 The spiral ligament is thought to play an important role in recycling potassium from the organ of Corti back into the scala media following hair cell transduction.22
An open and extensive fluid communication channel between the scala tympani and SGNs within Rosenthal's canal has important implications for the application of new-generation perimodiolar electrode arrays and cell- or drug-based therapies using a scala tympani route. The present results suggest that the OSL, separating the scala tympani from Rosenthal's canal, is likely to be far more electrically transparent than what has previously been thought. Ifukube and White23 used bone segments obtained from human cadaver cochleas to demonstrate that perforations in bone result in significantly lower resistance. While these authors compared bone samples with few pores with those containing numerous very large pores (100-200 µm; presumably taken from the medial wall of Rosenthal's canal), it is reasonable to also expect a low impedance pathway associated with the thin, highly porous OSL. Low impedance pathways will have an important influence on current flow in the implanted cochlea. Provided the patencies of the canaliculi are maintained following cochlear implantation, perimodiolar electrode arrays would be expected to produce relatively localized current spread.24
Any advantage the porous nature the OSL has may be lost in cases of cochlear pathologic conditions resulting in extensive fibrous tissue and/or new bone growth. For example, pathologic conditions such as bacterial or viral meningitis or electrode insertion trauma are likely to result in extensive new bone growth that would presumably obliterate the canaliculi perforantes. Such a pathological response may dramatically alter the patterns of current spread in the implanted cochlea.
The canaliculi of the OSL also provide a potential route for the application of pharmacological agents introduced into the scala tympani (either via direct infusion or via diffusion across the round window) for the protection of SGNs or hair cells during or following otologic insult. These bony pores may also provide a similar route for the application of cell-based therapies in the deafened cochlea. The potential to use stem cells or other cell lines to regenerate hair cells and/or SGNs is most appealing; moreover, many cell types are known to migrate in vivo.25,26 The canaliculi perforantes may, therefore, play an important role in allowing both pharmacological agents and exogenous cells to gain access to their target sites.
As noted previously, the patency of this network of pores is likely to be seriously compromised in some disease processes such as meningitis. Their extensive obliteration may result in only limited therapeutic advantage for pharmacological or cell-based therapies directly targeting SGNs or the organ of Corti via the scala tympani. Clearly, more research is needed to study in detail the role of the canaliculi in pharmacological or cell-based therapy targeting these cells. The establishment of a suitable experimental animal model of the human canalicular structure would be an important first step. Finally, the delivery of pharmacological agents via an electrode array27 may also be compromised by the formation of an extensive fibrous tissue capsule around the array. Research aimed at developing appropriate surgical delivery of these agents together with techniques designed to minimize fibrous tissue and osteogenesis after implantation is also warranted.
Corresponding author and reprints: Robert K. Shepherd, PhD, The Bionic Ear Institute, Department of Otolaryngology, 32 Gisborne St, East Melbourne, Victoria 3002, Australia (e-mail: firstname.lastname@example.org).
Submitted for publication July 15, 2003; final revision received December 12, 2003; accepted January 5, 2004.
This work was funded in part by grants from the Royal Victorian Eye and Ear Hospital, East Melbourne, Australia, the Garnett Passe and Rodney Williams Memorial Foundation, Parkville, Australia, and the National Institute on Deafness and Other Communication Disorders (contracts NIH-N01-DC-0-2109 and NIH-N01-DC-3-1005), Bethesda, Md.
This study was presented at the Australian Society of Otolaryngology Head & Neck Surgery 50th Annual Scientific Meeting, March 26-31, 2000, Melbourne, Australia; and the 2001 Conference on Implantable Auditory Prostheses, August 19-25, 2001, Pacific Grove, Calif.
We are grateful to Ivica Grkovic, PhD (Department of Anatomy and Cell Biology, University of Melbourne), who provided important training and advice on scanning electron microscopy techniques, and Michael Tykocinski, MD, and Rob Briggs, FRACS (Department of Otolaryngology, University of Melbourne), who provided helpful surgical advice during the study.
Create a personal account or sign in to: