Final Report Summary - NEUEAR (Neurotrophic Cochlear Implant for Severe Hearing Loss)
Executive Summary:
Sixteen percent of adult Europeans suffer from hearing loss, great enough to adversely affect their daily life. Over the age of 80, 50% of the population is suffering from hearing loss. A large portion of this population is affected by sensorineural hearing loss (SNHL), the consequence of a progressive degeneration of the primary auditory neurons (ANs), the afferent neurons of the cochlea. The only current therapeutic intervention for these patients is the use of a cochlear implant (CI), a neural prosthesis designed to directly electrically stimulate the ANs. Despite improvements in CI technology, the hearing experience with these devices remains far from normal, with many patients reporting difficulties discriminating speech in noise and poor perception of temporally encoded sounds such as music.
During the course of the grant, NeuEar partners have designed a neurotrophic cochlear implant – a novel neural prosthesis - to provide both electric auditory cues and regenerative neurotrophic factor(s) to severe-profoundly deaf patients. As the ongoing degeneration of ANs that occurs over time is a limiting factor in current cochlear implant efficacy, the ANs have been the target cells in the implant design strategy. The project aims were to develop an encapsulated cell (EC) therapy device capable of long-term intracochlear delivery of neurotrophic factors to prevent the degeneration of ANs, and to further combine this new EC device with a cochlear electrode implant for joint implantation in the cochlea.
As a first step, several clones based on a human retinal pigment epithelial cell line were developed to express high levels of either Glial cell-line Derived Neurotrophic Factor (GDNF) or Brain-Derived Neurotrophic Factor (BDNF). To test the clones in a relevant animal deafness model, a novel encapsulated cell device was engineered to allow for cochlear implantation in guinea pig and cat. In a one month (short term) guinea pig study, we showed that the encapsulated cells expressing either GDNF or BDNF successfully protected the animals from normally deafening, injected doses of neomycin. These very encouraging proof-of-concept results show that intracochlear delivery of neurotrophic factor from an encapsulated cell device can indeed provide a positive effect to diminish hearing loss.
In a following long-term, 6-month cat study, encapsulated cell devices were implanted together with custom-made cochlear implants designed by the consortium. Good results on safety and the lack of adverse effects in the cats were obtained, but the cochlear implants were generally broken over the six months due to the natural movements of the cats. Furthermore, encapsulated cell devices implanted over the same time were found to be rejected by the cats, leading to their loss of function. Compared to the good results obtained in guinea pigs, this indicates that the cat deafness model could be difficult to work with for chronic cochlear implantations.
In parallel to the in vivo studies, several design concepts for a final clinical product were made, resulting in functional prototypes that can be employed in coming clinical studies. These designs are currently being patented and will, based on the very encouraging NeuEar data, be used in a clinical Phase I trial planned to start shortly after the end of the NeuEar project.
Project Context and Objectives:
Project context
Hearing loss (HL) is the most prevalent sensorineural disorder in developed countries and the most common birth defect. One of every 500 newborns is affected with bilateral congenital sensorineural hearing loss (SNHL) ≥40dBHL. This number rises to 2.7 per 1000 before the age of 5 and 3.5 per 1000 during adolescence (Morton, 1991). The SNHL typically occurs following damage and loss of hair cells - the sensory cells in the cochlea of the inner ear - which, in response to sound, convert the mechanical vibrations into nerve impulses in the primary auditory neurons (ANs). Widespread hair cell loss results in severe-to-profound SNHL (Nadol et al., 1997). The only therapeutic intervention for these patients is the use of a cochlear implant (CI), a neural prosthesis designed to directly electrically stimulate the ANs. The CI system consists of an external unit and an implanted, internal unit. The behind-the-ear external processor with ear hook (1) and a battery case (2) uses a microphone to pick up sound, converts the sound into a digital signal, processes and encodes the digital signal into a radio frequency (RF) signal, and send it to the antenna inside a headpiece (3). The headpiece is held in place by a magnet attracted to an internal receiver (4) placed under the skin behind the ear. A hermetically sealed stimulator (5) contains active electronic circuits that derive power from the RF signal, decode the signal, convert it into electric currents, and send them along wires (6) to electrodes (7). The electrodes are assembled on the silicone rod (electrode array) and inserted into the cochlea, where they stimulate the auditory nerve (8). The nerve conveys the signal to the central nervous system, where the electrical impulses are interpreted as sound.
Despite improvements in CI technology, the hearing experience with these devices remains far from normal, with many patients reporting difficulties discriminating speech in noise, and poor perception of temporally encoded sounds such as music. Part of the inefficiency of cochlear implants is related to a poor interface with and progressive degeneration of the ANs, ultimately leading to significant neuronal loss after long periods of deafness. In addition, experimental animal studies indicate that on-going AN degeneration compromises the efficacy of the cochlear implant.
The loss of endogenous neurotrophic factors normally expressed by hair cells, such as brain derived neurotrophic factor (BDNF) and neurotrophin-3 (NT3), initiates this AN degeneration. Numerous studies have demonstrated that intracochlear infusion of these neurotrophins can support AN survival in animal models of deafness. Furthermore, when combined with chronic electrical stimulation via a cochlear electrode, exogenous neurotrophin treatment results in significantly enhanced AN survival compared to neurotrophin treatment alone (Shepherd et al. 2005). The novel EC implant to be developed in this project takes advantage of cell-based de novo synthesis of the therapeutic neurotrophin and allows for the chronic production of bioactive protein. Hence, the encapsulated cell device aims to combine the advantages of gene therapy with that of the safety of a removable device, attached to the cochlear electrode itself or as a separate implant.
The EC technology consists of an elongated device that contains a genetically modified human cell line enclosed behind a semi-permeable hollow fibre membrane that allows for the influx of nutrients and the outflow of internally produced therapeutic factor(s), but does not allow for the direct contact of the cell line and the host tissue.
The implant provides for long-term factor secretion of de novo synthesized protein while allowing for its retrieval. In addition, the permselective membrane prevents immune rejection of the allogeneic cell line, allowing for the same parental cell line bank to be used to produce the various therapeutic cell lines in all products and individuals. The long-term stability (1 year) of this delivery system in the brain has been demonstrated with an EC therapy product secreting nerve growth factor (NGF). This product is currently in clinical trials for the treatment of the cholinergic neuronal degeneration in patients with Alzheimer's disease. A total of 10 patients have been implanted with up to four NsGene clinical devices per patient and the technology has thus been through technical, manufacturing, safety, functional, and regulatory proof of concept. With the proposed research and development with the collaborators and work packages in this project, it is anticipated that the NsGene encapsulated cell technology can be modified and developed for use in the inner ear within the proposed three-year timeframe.
Project objectives
The NeuEar project aimed to develop genetically modified cells for encapsulation and long-term expression of selected neurotrophin(s) in a medical device implant prototype that can be combined with a cochlear implant. The NeuEar partners have also evaluated and optimized safety and functional effects in in vitro and in vivo assays and evaluated the long-term safety and efficacy in a guinea pig and cat animal models.
The final goal was to have an implant prototype and functional data in suitable animal models available that support a continuation towards clinical testing after the NeuEar project has been finished in the proposed 3-year period.
The specific objectives of the NeuEar project were to:
(i) develop genetically modified cells for encapsulation and long-term (>6 months) over-expression of selected neurotrophin(s) in vivo
(ii) develop EC implant prototypes that can be combined with a cochlear implant
(iii) evaluate and optimize safety and functional effects in in vitro and in vivo assays
(iv) evaluate the long-term safety and efficacy in a large animal model of a clinically relevant implant
(v) implement an efficient dissemination structure, including a patent strategy to enable partnering and fund-raising for further clinical development, regulatory approval, commercialization, and marketing.
References
• Morton NE. Genetic epidemiology of hearing impairment. Ann NY Acad Sci 1991;630:16-31
• Nadol JB Jr. 1997. Patterns of neural degeneration in the human cochlea and auditory nerve: implications for cochlear implantation. Otolaryngol Head Neck Surg. 117:220-8.
• Shepherd RK, Coco A, Epp SB, Crook JM. (2005) Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J Comp Neurol. 486(2):145-58.
• Zeng, F.; Rebscher, S.; Harrison, W.; Sun, X. & Feng, H. (2008) Cochlear Implants:System Design, Integration and Evaluation. IEEE Rev Biomed Eng, 1, 115-142.
Project Results:
In the first half of the NeuEar project, a particular emphasis was given for the selection of the most appropriate cell lines and neurotrophic factors to be used in the EC device. In this last half of the project, characterization of the developed cell lines was further continued by Encapsulated Cell (EC) device implantation in both guinea pig and cat animal deafness models to identify a candidate neurotrophic factor and cell line for treating sensorineural hearing loss. First results came from guinea pigs where both cell lines tested – one secreting GDNF and one BDNF – were shown to protect spiral ganglion neurons from the deleterious effects of neomycin injection over the course of one month. Compared to untransfected ARPE-19 cells (not expressing exogenous neurotrophic factors), the GDNF #125 cell line showed a significantly smaller increase in eABR threshold with a very narrow standard deviation among samples. The BDNF #202 cell line also caused eABR thresholds to remain low, but variation in eABR thresholds was larger than for the GDNF line. Based on these very encouraging data, the GDNF #125 cell line was chosen as candidate cell line for the remainder of the NeuEar project with BDNF #202 as backup cell line. The CDNF #605 line that was a contender during the first project period was finally abandoned, as several attempts of redesigning the amino acid sequence to increase expression ultimately failed.
As result of the short-term guinea pig data, the long-term, chronic cat study was designed to focus on GDNF #125, with a few pilot cats implanted with the BDNF alternative. Where the short term guinea pig study made use of already well-established cochlear implant electrodes, CIs for the cat study had to be custom made by Partner 3 in collaboration with Partner 4. It quickly became evident that designing a durable device able to maintain its function over six months in the very agile cat was to be a problem. Most electrodes failed shortly after implantation and it took several design iterations and implantation procedure optimizations to increase CI longevity sufficiently, which in turn decreased the number of functioning animals available for analysis at the study end. Most encapsulated cell devices filled with GDNF #125 cells unfortunately did not show any neurotrophic factor expression after 6 months in vivo. The reason for this is most likely a strong fibrotic reaction to the implanted devices, as could be observed from the histologic analysis of explanted cochleae. Combined with the electrode challenges, we were therefore not able to obtain as robust data in the cat as in the guinea pig. Nonetheless, SGN protection was still evident in the experimental group combining GDNF device implantation with electrical CI stimulation (ES). GDNF devices without ES did not show a protective effect, raising the concern that GDNF treatment in itself did not result in a protective effect. However, the number of animals in each group was too low and the electrode variability too high to make a firm conclusion.
In addition to the cat CI design, concept EC and CI devices for human implantation were successfully designed and developed. Two approaches were chosen, one where the EC and CI were fused to each other allowing for a single-implantation procedure, and one approach where the CI and EC were implanted separately. Several iterations of both concepts were tested in cochlear plastic models as well as in human temporal bones. While both concepts were feasible, the advantages associated with the separated approach, in particular with respect to flexibility and little need for CI modification, were found to outweigh the surgically simpler fusion device. For use with the separate EC implantation, an implantation port concept named EPS (Extension of Perilymphatic Space) was developed. Several EPS design iterations were successfully tested in human temporal bone.
Details of the major achievements in each scientific work package are summarised below.
WORK PACKAGE 1 – SELECTION OF NEUROTROPHIC FACTORS AND GENERATION OF CELL LINES
Cell line development (Wahlberg, Tornøe, Ulfendahl, Lenarz, Scheper)
The first task of the NeuEar Work Package 1 concerned the development of cell lines to be used in Partner 1’s (NsGene, Denmark) EC Biodelivery technology for cochlear application. The project started out with a plan focusing on developing ARPE-19 based cell lines expressing the neurotrophic factors brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and Cometin, the latter being a neurotrophic factor discovered recently by Partner 1. NT-3 and Cometin was later replaced with glial cell-line derived neurotrophic factor (GDNF) and cerebral dopamine neurotrophic factor (CDNF). GDNF was chosen for its well-documented trophic effect on a wide range of neurons, including auditory neurons, and the novel factor CDNF was selected as a scientifically interesting and exploratory molecule to test due to its possible effect on auditory neurons combined with very few reports on the application of CDNF for hearing indications thus far.
To be able to generate high-producing cell lines expressing the three neurotrophic factors of choice, plasmid vectors based on Partner 1’s proprietary Sleeping Beauty (SB) transposase expression system were generated. Briefly, the SB system allows for high-level genomic integration in the recipient cells mediated by a co-transfected transposase, leading to strong and stable expression of the introduced transgenes. The completed vectors were tested transiently for their ability to mediate factor expression from ARPE-19 cells, followed by generation of stable cell clones by antibiotic selection. All plasmid preparations used for the cell clone generation was made under GMP at Plasmid Factory, Germany to ensure that no adventitious agents were introduced.
The stable transfections resulted in a large set of clones to be characterized for their neurotrophic factor production levels. While expression of GDNF and BDNF was high (up to 12 µg factor / 106 cells / 24 h), CDNF expression from isolated clones was about 50 times lower, on average. CDNF transfection was repeated several times with similar results. As the CDNF levels obtained from the ARPE-19 clones could not reach physiologically relevant levels, the project focus was directed to GDNF and BDNF clones. Here, GDNF clone #125 and BDNF clone #202 were chosen as candidate clones for further studies.
Cell line selection (Wahlberg, Tornøe, Ulfendahl, Lenarz, Scheper)
After having identified the high-producing clones GDNF #125 and BDNF #202, further characterization was needed in order to qualify the clones for the later in vivo experiments in deafened guinea pig and cat. From the ELISA data we knew that high levels of immunogenic protein was released from the cells, but further testing of the biological activity was warranted. In theory, truncated protein variants expressed from the ARPE-19 clones could be ELISA positive while having lost their biological effects.
Furthermore, the in vitro factor release level is but one of several important properties of a clinically relevant cell line. Equally important, a suitable cell line must show good growth properties and be able to survive in the more restrictive in vivo environment after implantation. This was tested by striatal implantations in Göttingen minipigs and rats. Both clones showed sustained factor production after up to three months implantation of the EC device. Immunohistochemistry of rat striatum implanted with GDNF #125 showed that GDNF produced from the devices did readily diffuse into the target tissue. Furthermore, the biological activity of GDNF was assessed by staining for tyrosine hydroxylase (TH) which is commonly known to be upregulated by GDNF. An increased TH staining level in the implanted side compared to the control side confirmed that the GDNF secreted from the implanted EC device was indeed active.
Due to technical issues at Partner 2, the in vitro testing of BDNF #202 was delayed. To carry on testing the clone anyway, it was included in a guinea pig evaluating the AN-protective effects of the encapsulated cells. Here, both the GDNF and BDNF clone showed good protective effects compared to control cell not expressing measurable amounts of neurotrophin.
In conclusion, both the GDNF and BDNF clones produce such high levels of neurotrophin that therapeutically relevant release levels can be achieved. For the CDNF clones, however, we have not been able to reach this goal. Despite a large number of attempts, the best of the isolated CDNF clones have expressed almost two orders of magnitude less factor than their GDNF and BDNF counterparts. Consequently, we have moving forward with GDNF and BDNF without CDNF. Solving the CDNF expression issue calls for redesign of the entire expression system to accommodate this factor. Due to the high relevance and novelty of CDNF application for deafness, Partner 1 and 3 continued to pursue this goal and finally showed that therapeutically relevant doses of CDNF were indeed possible to achieve in the in vitro SGN assay.
WORK PACKAGE 2 – DEVICE DESIGN AND PROTOTYPING
Design and manufacturing procedures for animal and clinical devices (Wahlberg, Tornøe, Ulfendahl, Jolly, Mistrik)
All materials employed for device construction are made from clinical grade materials with excellent biocompatibility to minimize tissue reactions after implantation. EC biodelivery devices are made from hollow fiber polysulfone membrane. The membrane dimensions are tailor-made to accommodate the application. The brain device currently employed by Partner 1 (NsGene, Denmark) is made from a membrane developed by NsGene, having an outer diameter of 700 μm and an inner diameter of 550 μm. The molecular weight cut-off is 280 kDa. This membrane configuration will be used for developing the human clinical prototypes in the NeuEar project.
For guinea pig and cat cochlear application, Partner 1 has developed a smaller-diameter polysulfone membrane with a similar molecular weight cut-off. The outer diameter is 400 μm and the inner diameter is 300 μm, resulting in a thinner wall with better diffusion properties. The length of the guinea pig device is 3,5 mm long. For cat cochlear application, a 4 mm long device with the same diameter is used. As scaffold material, a multi-filament yarn material from Swicofil, Switzerland is used, providing the ARPE-19 cells with an extensive surface area on which to attach. Both ends of the device are sealed by medical grade, UV-polymerizable glue.
Cell-containing devices are maintained in serum free medium at 37°C, 5% CO2 until ready for implantation. Manufacturing, filling and sterilisation of EC devices is done entirely independently from that of the cochlear implants to provide a simple functional assembly of the two devices before surgery.
The cochlear implant for the clinical device is the standard, commercially available, CI system manufactured by Partner 3 (Med-El, Austria). Two design approaches for the clinical device have been decided. In the first approach, EC and CI devices are implanted independently. The CI electrode array is implanted first, followed by the EC device. In the second scenario, the EC device is assembled on a modified electrode array of the CI system by a surgeon prior to implantation using a simple clipping mechanism, and both devices are implanted simultaneously. This design is still in the prototype phase.
Neurotrophic cochlear implant for guinea pig implantation (Wahlberg, Tornøe, Jolly, Mistrik)
Guinea pig specifications
The straight part of the guinea pig cochlea in its basal turn is about 6 mm long. Theoretically, it can accommodate a device with a length of up to 5 mm with an outer diameter of 0.6 mm. The electrodes for electrical stimulation are made from 90% platinum and 10 % iridium and are coated with Teflon for insulation. The diameter of the stimulus electrode is 0.075 mm and for the ground electrode the diameter is 0.125 mm. For this particular use in the combination with the EC device, a small ball (0.2-0.3 mm) is burned on the stimulus electrode by using oxygen gas to get a bigger surface when stimulating the auditory nerve. The stimulus electrode is inserted 2-2.5 mm through the round window into scala tympani, the other end of the electrode is stripped from Teflon and soldered to a percutaneous connector. The ground electrode is stripped from insulation approximately 20 mm at both ends, one end is placed along the wall in the middle ear cavity, and the other end is soldered to the same percutaneous connector as the stimulus electrode.
Device construction
A biodelivery device suitable for implantation in the guinea pig cochlea was constructed using the 400 μm membrane developed specifically for NeuEar project. The device is 3,5 mm long and can be placed in the guinea pig cochlea base without disrupting the organ tissue. An inert device handle has been developed to facilitate positioning in the cochlea. The active part of the device is filled with the neurotrophin-producing cell line to be ready for implantation.
Neurotrophic cochlear implant for cat implantation (Wahlberg, Tornøe, Jolly, Mistrik)
Cat specifications
The straight part of the cat cochlea in its basal turn is about 6 mm long and the diameter is about 2 mm. It can accommodate two 5 mm long devices with outer diameter of 0.6 mm each, allowing for parallel implantations of the EC device and CI electrode array. The EC device cannot be bent due to the rigid nature of the polysulfone membrane used for device manufacturing, whereas the electrode array of the CI is made from silicone elastomer providing high flexibility. The EC device is 4 mm long and has an outer diameter of 400 μm. The electrode array for the CI is 8 mm long and the diameter is 0.3 mm at the tip and 0.6 mm at the end.
Device construction
The EC device with cat specification is constructed using manufacturing procedures described above. The CI electrode array was made from platinum and iridium wires coated with Teflon. Each wire is connected to a single contact in the array tip and embedded in silicone using a standard animal-grade aluminum mold. The diameter of the array is 0.3 mm at the tip and 0.6 mm at the end. The length of the array is 8 mm. The wires are 20 mm long. Each of them is connected to a 9-strand platinum wire, which is 150 mm long.
The device was manufactured in two variants. In the first design, the multi-stranded wires are connected to the pins on a transcutaneous connector. The Pulsar stimulator is attached to the second part of the connector. This variant is intended for the implantation of the electrode array, while the stimulator being carried out externally by animal in a backpack (transcutaneous approach). In the second design, the connector is omitted and the multi-strand wires are connected directly to the Pulsar stimulator. This design is intended for subcutaneous implantation of the entire electrode system, including the stimulator.
In the current EC device / CI design, the two devices are kept separate and the EC device is not assembled on the cat electrode array. This design has been adopted after initial tests in temporal bones from cat cadavers, where sequential implantation of the two devices proved to be more efficient. This corresponds to the first of the two clinical approaches. First, the electrode array is inserted through the opening in the round window, after which the EC device is introduced through the same opening in parallel to the electrode array.
Clinical prototypes of neurotrophic cochlear implant
Combined device design
For the construction of the prototype of the neurotropic implant a novel design of the electrode array was needed to be developed to accommodate also the EC device inside the scala tympani (ST) with a limited volume. This design carried a reduced number of electrode contacts, 12 instead of 24, which are organised in one line. In this design, each contact corresponds to a single electrode, and the surface is increased to 0.14 mm2. The planned diameter of this electrode array is 0.6 mm and 0.4 mm at the cochleostomy and tip, respectively. The reduced diameter from 1.3 mm to 0.6 mm is to allow a sufficient space in the basal region for the attachment of the EC device. The EC device was planned to be attached to the array at its basal end at close proximity to the cochleostomy, by a clipping mechanism. For biocompatibility reasons, such clips were planned to be made from silicone, the same material, from which the rest of the array was also made.
The planned clipping mechanism made from biocompatible silicone was also attached to the electrode array and its functionality tested. The extensive testing revealed that this type of the attachment is not optimal, because it increases significantly the diameter of the array in the area where the EC device is attached. This would be detrimental for the insertion, which must be maximally atraumatic. To overcome this problem, an alternative design was invented, based on the stent concept. The stent-like half-cylinder is integrated into the silicone rod to accommodate the EC device inside the electrode array without increasing significantly the diameter.
The first prototype of the scaffolding was made from a 50μm thick platinum sheet with laser-cut openings in it. The sheet was bent to semi-circular form and incorporated into the mould for electrode fabrication. The wires running to electrodes positioned in more apical electrodes are also visible. The implant was extensively tested using the plastic model of the human cochlea. Unfortunately, the repeated insertions revealed that the platinum-sheet based cage is too rigid for atraumatic insertion.
Therefore, the platinum-based half-cylinder was replaced with a polymer mesh tube made from polyimide. The final configuration is with the polymer cage assembled on the electrode array. The EC device is partly exposed. The cage was made in several versions of different length: 7 mm to accommodate the 5-mm long EC device and 5 mm to accommodate the 4-mm long EC device.
Mechanical and insertion tests using the plastic model of human cochlea and artificial temporal bone revealed that also in this configuration the device is relatively rigid and the insertion into a real temporal bone might not be sufficiently smooth to prevent any damage to cochlear structures required for the residual hearing preservation. Therefore, in the final prototyping step the cage was replaced by a more flexible cage, made from the same material, but with the central part cut out to reduce rigidity.
The implantability of both prototypes was tested in cadaveric temporal bones by partner 4. Full insertion was achieved with both prototypes, but only after the electrode array segment with the first 4 electrodes was cut away. Consequently, the maximal possible length of the electrode array was determined to be 24 mm, and not 28 mm of the original design. Therefore, the electrode array was shortened to 16 mm, to provide a sufficient margin for any variation in the cochlear length, and electrodes were redistributed accordingly along the length of the array.
Furthermore, the histological analysis of the cochlea implanted with any of the two designs revealed the full-cage design as less traumatic, even if also in this case some bending of the basilar membrane was observed. In summary, the testing indicated that using the shorter electrode array, Flex16 design, with integrated the full polymeric cage, would be the best design for the combined device, in which the EC device and CI electrode array are integrated into a single implantable device.
Finally, the second prototype of the combined device design was manufactured with a polymer cage of reduced size to accommodate the smaller EC device used in animal experiments in WP3 and WP4 (diameter 0.4 mm, length 4 mm). In principle, such EC device could be also clinically relevant, if the therapeutically agents produced in reduced quantity still have protective/regenerative effect on the target tissue. The dimensions of the cage are: diameter 0.6 mm and length 7,3 mm. The overall diameter of the implant is 1.1 mm, which is smaller than 1.3 mm for the Standard electrode array design (at the level of the cochleostomy entry) and the implantation of this version should be very atraumatic.
Separated device design
The prototyping of the second design, with the EC device and CI electrode array implanted into the cochlea separately, was based on the feasibility testing in cadaveric temporal bone carried out in the first reporting period. After demonstrating the feasibility of parallel independent implantation of both the EC device and CI electrode array into the scala tympani of the cochlea (or of the EC device alone), the question of the independent withdrawal of the EC device, while keeping the CI electrode array in place, was investigated. For that reason, a harbouring port, made from titanium for biocompatibility, and permanently fixed in the temporal bone opening drilled by a surgeon next to the round window was developed.
The thickness of the promontory region of the temporal bone in the target area was measured from the CT scan of an implanted cochlea and estimated to be between 1.6 and 1.7 mm. Therefore it is anticipated that 2 mm long opening must be drilled through the bone to reach the scala tympani. The port was therefore designed accordingly. The specifications (2.2 mm length, 1.3 mm diameter) and the first prototype made from titanium. The port has an extra handle for correct positioning of the sliding part in the bone. The insertion of the port was first tested in a plastic model of human cochlea. It demonstrates that there is enough space to accommodate both devices. Next, the port was tested in the temporal bone. The insertion was done by partner 4. The cochleostomy for the port was drilled next to the round window, the port was fixed in it and the EC device (Length 7 mm) inserted. The Flex16 electrode array was inserted independently through the round window. The consequent histological analysis of implanted temporal bone revealed that 1) the cochleostomy needed for a port of these dimensions was too large (too traumatic) and 2) that the handle was not needed for the correct orientation of the port in the bone. On the contrary, it made the insertion too cumbersome. Instead the screw-thread should be added for better fixation in bone.
These observations were taken into account in the design of the second prototype of the port. The inner diameter was reduced (1.17 mm), the lower part, fixed in the bone, made longer (1.45 mm) and a screw-thread added. The temporal bone insertion test revealed that the dimensions are still unnecessarily large and that the helical configuration of the screw-thread is needed for firmer fixation.
The final prototype (version III) with the helical screw-thread and even smaller inner diameter (0.8 mm). It bears an incision in its upper surface for mounting the port into the bone using a standard microsurgical screwdriver. A mark (2 round spots) is also included to facilitate the correct orientation.
The EC device bears in this final configuration a concept handle made from PEEK for easier handling and hermetical sealing of the port. Such handle could be used also for the device retrieval, if needed. In addition to the port design for the EC device with 0.7 mm diameter, the second version for the smaller EC device (diameter 0.4 mm), used in animal experiments, was also produced to demonstrate the feasibility of the manufacturing, if such smaller device would be used clinically.
Device design
During the second half of the project partners 1 and 3 have closely collaborated on the development and testing of two versions of the clinical neurotrophic implant. In the first configuration the EC device and the CI electrode array were integrated together to form a single implantable device (Combined device design). In the second configuration the designs were modified to facilitate independent but parallel implantation of the two devices into the cochlea (Separated device design). The feasibility of implantation was tested in cadaveric temporal bones by partner 4.
In its final design, the first configuration features the polymeric mesh permanently attached to the CI electrode array of reduced diameter (0.6 mm) and length (16 mm) to accommodate the EC device. Two versions were made to fit a possibly different diameter of EC device (0.7 and 0.4 mm). For the second configuration, the port permanently fixed in the promontory region of the temporal bone was designed, to facilitate independent insertion and retrieval, if necessary, of the EC device through cochleostomy. The CI electrode array is inserted through the round window in this case. Again, also this configuration was prototyped in two versions to accommodate two possible diameters of the EC device.
WORK PACKAGE 3 – FUNCTIONAL TESTING IN VITRO AND IN VIVO
Cellular effects of factors released by encapsulated cells on inner ear tissue (Wahlberg, Tornøe, Ulfendahl, Dash-Wagh, Jolly, Mistrik, Lenarz, Scheper, Hoffmeister, Kristof)
The objective of this task is to demonstrate that factors released by selected clones give positive neuronal effects. This has been addressed by testing the survival of auditory cells exposed to factors produced by cell clones used in encapsulated cell implants. The experiments were performed using spiral ganglion cell cultures to analyse the neuroprotective effect of ARPE-19/ NT cells producing BDNF, GDNF or CDNF compared to control and conditioned medium derived from ARPE-19 cells.
The effects of BDNF, GDNF, and CDNF secreted by the modified ARPE-19 cells on the survival and growth of auditory neurons was investigated using organotypic cultures. The in vitro cultures from the spiral ganglion provide a system that allows for a faster and easier screening of the protective effects of the different neurotrophic factors (NTs). These cultures are physiologically closer to in vivo than the dissociated cells. The effects of NTs will be evaluated in terms of the number surviving neurons, and neurite outgrowth and lengths using immunocytochemistry, microscopy and image analysis.
Survival and growth of auditory spiral ganglion (SG) neurons exposed to factors released by encapsulated cell (EC) devices in co-cultures
Using in vitro spiral ganglion explant cultures were used to analyse the neuroprotective effects of ARPE-19/NT cells producing BDNF, GDNF or CDNF. The release of NTs from individual devices was tested using enzyme-linked immunosorbent assay (ELISA) prior to experiments and in the presence of auditory neurons. The levels of CDNF release were very low compared to the production of BDNF and GDNF devices. The ARPE-19/GDNF and CDNF devices showed less variation in NT production compared to BDNF devices. Explants of the spiral ganglion were co-cultured in 4-well chambers together with NT-producing EC devices, placed at a distance from the tissue of about 2.5-3 mm. After 4 days of incubation, the cultures were fixed and immunolabeled with antibodies against neurofilament and tyrosine kinase receptor B (TrkB). Using fluorescence microscopy the size of explants was measured and the number of neurofilament 200 and TrkB positive neurons was estimated as well as the number and length of the neurites.
Consistent with previous studies, co-culture of neonatal SG explants with BDNF-producing EC devices resulted in a significant increase in the number of neurons in SG explants compare to control, CDNF and GDNF. In addition, the number of TrkB positive neurons were also higher compared to other conditions. To our knowledge, this is the first report of a neuro-protecive effect of CDNF on the SG neurons in vitro. CDNF significantly enhanced the survival of SG neurons (both NF200 and TrkB positive) as compare to control and GDNF. Although low concentrations of CDNF were produced by the EC devices, the effects were similar to that of BDNF devices.
Co-culture with GDNF-producing EC devices slightly but not significantly increased the neuronal survival in the SG explants, which is in contrast to previously published in vivo reports. The difference could be attributed to the different model systems. However, our data is in line with effects reported by other groups on dissociated neonatal or adult SG neurons where GDNF showed no significant neuro-protective effect.
The neuritogenic effects of the released NTs were analyzed by quantifying the number of neurites extending from each explant. The untreated explants showed the fewest neurites while explants co-cultured with BDNF producing EC devices exhibited the highest number of neurites. CDNF also significantly increased the number of neurites extending from each explant. GDNF-releasing EC devices, despite producing amounts of GDNF (> 100 ng/ ml/ 24 hr) expected to have a neuritogenic effect, did not promote neurite outgrowth in the observed SG explants. The length of neurites was strongly increased in presence of BDNF- and CDNF-producing EC devices while GDNF showed a marginal and non-significant increase.
In conclusion, the results indicate that the NT-EC devices produce similar amounts of NT that have previously been reported to induce survival and neurite outgrowth from the SG neuron both in vivo and in vitro. BDNF and CDNF producing EC devices displayed strong neuroprotective and neurogenic activity on SG neurons in the explant cultures. In contrast, GDNF EC devices showed no influence on either survival or neurite outgrowth of SG neurons. A novel finding of this study is the influence of CDNF on SG neurons.
Further studies on CDNF to follow up novel findings in Task 1
As reported above, it was observed that the CDNF released from the encapsulated cell device (ECD) promoted neuronal survival and neurite outgrowth in the spiral ganglion explant. Interestingly, despite the low concentration of CDNF released by the ECD (compared to BDNF released by ECD), the strong neurogenic effects suggested a higher efficiency of CDNF. To induce these effects in spiral ganglion, the neurons should express the receptor of CDNF. Unfortunately, there are no reports about the expression of CDNF in inner ear tissue, and the receptor for CDNF is yet unknown. We performed PCR and immunohistochemical studies to investigate the expression of mRNA and protein CDNF in the postnatal rat cochlea and spiral ganglion explants. Real-time PCR results showed that the transcript of CDNF is expressed in the cochlea during early development. However, expression in the adult is higher compared to the early postnatal ages. The immunohistological studies reveal the expression of CDNF protein in the spiral ganglion neurons particularly in the neurites; both in the P2 cochlea and in the explant cultures prepared from the P5 rats. There was no or very marginal expression of CDNF in the cell bodies of the spiral ganglion neurons in the explants. Similarly, the hair cells exhibited no CDNF expression in the organ of Corti. High expression was observed in the neuronal terminals and in the postsynaptic zone under the hair cells. This expression patterns suggest that spiral ganglion neurons express the CDNF and possibly its receptor, thus, can respond to the exogenous CDNF.
Effects of electrical stimulation on factor-releasing encapsulated cells and spiral ganglion neurons in vitro (Wahlberg, Tornøe, Ulfendahl, Dash-Wagh, Jolly, Mistrik, Lenarz, Scheper, Hoffmeister, Kristof)
During the first project month we tried to establish the electrical stimulation of ECs using the Zahner System which is, for single electrical stimulation of seeded cells, established in our lab. Unfortunately we failed and needed to establish an alternative method for EC stimulation. With help from partner 2, MED-EL, who provided a DIB-Box we were able to build up a new stimulation system which now allows us to electrically stimulate 4 wells in parallel using the same stimulation parameters.
In December 2013 the first electrodes with a “clip-on” system for ECs build from MED-EL arrived in our lab and we were able to perform first stimulation experiments using ECs with encapsulated BDNF and GDNF producing ARPE cells. Briefly, ECs were clipped on the electrode, electrode and EC-device were placed in a chamber slide and fully covered with medium. An initial impedance measurement was performed. Afterwards the electrical stimulation (1000pps, 1200µA) started for 12 hours. Every four hours the supernatant of stimulated ECs was collected, frozen and after finalizing the experiment all samples were sent to NsGene for evaluation of neurotrophic factor concentration by ELISA.
The highest factor production was observed in GDNF-producing ARPE-19 cells. Electrode impedance measurements were performed on each time point before medium extraction and no stimulation related changes in impedances where observed for any of the tested cell lines at any of the measurement time points. Based on factor production, these results indicate that the GDNF-producing clone is the most promising candidate.
The effects of surgically implanting an encapsulated cell device into the inner ear in an in vivo animal model (experimentally deafened guinea pigs) (Malte Hoffmeister, Alexander Kristof, Verena Scheper, Wiebke Konerding, Jens Tornøe, Jenny Ekberg, Briannan Bintz, Chris Thanos, Pavel Mistrik, Anandhan Dhanasingh)
To explore possible functional benefits, a stimulus electrode is implanted at the same time in order to electrically evoke an electrophysiological response from the remaining auditory neurons (spiral ganglion neurons) during several weeks following deafness and implantation.
Guinea pigs with normal hearing were experimentally deafened by transtympanic injections of the ototoxic drug neomycin. After 3 weeks, when significant hearing loss was demonstrated by recording acoustically-evoked auditory brainstem responses, the animals were implanted with one of four different types of EC devices (length: 4mm, diameter: 0.4mm):
1. ARPE-19/GDNF
2. ARPE-19/BDNF
3. Untransfected ARPE-19
4. Empty devices (not containing cells)
The devices were placed into the scala tympani in the basal turn of the cochlea. At the same time, an electrode mimicking a cochlear implant (CI) was also implanted through the round window into the scala tympani. The electrode device was only used for eliciting electrically-evoked auditory brainstem responses (eABRs) not for electrical stimulation. To determine changes in SGN electrical responsiveness, eABRs were measured weekly for 4 weeks. After the last eABR measurement, EC devices were retrieved and neurotrophic factor release levels were measured. The animals were sacrificed and the cochleae collected.
After decalcification, the cochleae were prepared for structural analysis in order to estimate the density of remaining SGN. Also the mean sectional area of the SGN was measured.
Electrically evoked auditory brainstem response (eABR)
To investigate if EC device delivery of neurotrophic factors improved the survival of the SGNs, electrical responsiveness was measured weekly during the 4 weeks that the treatment lasted. The eABR results displayed a significant difference between the two groups treated with neurotrophic factors compared to the two negative control groups. Between the two treated groups, the ARPE-19/GDNF group shows significantly lower thresholds than the ARPE-19/BDNF group. There is a considerable variation in the eABR results within the ARPE-19/BDNF group compared to the ARPE-19/GDNF group. It is possible that this due to the fact that the neurotrophic factor release from the ARPE-19/BDNF EC-devices was considerably lower than the neurotrophic factor release from the ARPE-19/GDNF EC-devices.
Morphology
It has been hypothesized that the electrical responsiveness of the inner ear is closely related to the number of remaining SGNs in the Rosenthal´s canal. To investigate the amount of remaining SGNs after 4 weeks of EC device treatment, the SGN density in the Rosenthal´s canal was calculated using the Zeiss ZEN software Image. At the same time SGN areas were measured. The group that received ARPE-19/GDNF treatment had a significantly larger amount of remaining SGNs than the groups that received ARPE-19/BDNF, ARPE-19 or the empty devices.
WORK PACKAGE 4 – LONG-TERM SAFETY AND FUNCTIONALITY IN LARGE ANIMAL MODELS
Generating deafened animal model (Verena Scheper, Wiebke Konerding, Jana Schwieger)
All relevant SOPs and protocols for in vivo cat experiments were prepared and rapidly received an ethical approval. Within thr first project year, we received the permission to work on n=9 cats for the pilot trial. These animals have been deafened and implantated with an EC device as well as a cochlear-implant.
Pilot in vivo tests for proof of feasibility of the novel biodelivery devices (Verena Scheper, Wiebke Konerding, Jana Schwieger, Jens Tornøe, Jenny Ekberg, Chris Thanos, Briannan Bintz, Pavel Mistrik, Anandhan Dhanasingh, Jochen Tillein, Raphael Kiran)
Within the first project year we applied for the ethical approval for the pilot trial and received the permission for n=9 cats to be chronically deafened and implanted as pilot animals. Next to this, we performed temporal bone studies on EC improvement. Cat cadavers were co-implanted with EC devices designed for the cat cochlea by Partner 1 (NsGene, Denmark) and cat CIs for evaluation of optimal EC design. Cadaver experiments were performed with Dummy ECs of distinct diameters which were co-implanted together with CIs into cat scala tympani via the round window. Based on those experiments, we found the optimal design of the cat EC device to be 4 mm length and 0.4 mm diameter with a tether on one end for better handling.
The experimental animals were chronically deafened using daily subcutaneous injections of the aminoglycoside neomycin sulphate (60 mg/kg BW s.c.). So far, six animals have been included into the pilot trial. The first cat was deafened starting from postnatal day 14 after an evaluation of normal hearing status via AABR. Hearing status was assessed after 14 days (and 19 and 24 if necessary) of daily injections. As this cat did not show complete hearing loss after the maximal duration of chronic deafening, we decided, based on literature survey and discussions with other working groups, to switch to a neonatal deafening procedure starting on postnatal day 1 to 3. All of the following five animals lost their hearing ability within 24 days after birth/start of deafening procedure. None of the deafened animals showed clinically relevant signs of kidney failure, which can be a negative effect of aminoglycoside treatment. From the first three pilot cats blood samples were taken on the day of implantation and analyzed for kidney parameters. No relevant changes have been observed. The kidneys of the first three pilot animals were investigated by a pathologist for kidney anomalies and the histology did not show any pathological changes. Therefore we conclude that the neonatal deafening procedure is leading to a robust deafening which allows for further processing within this trial and that the deafening does not have relevant negative effects which may lead to a premature termination of experimental animals.
The five pilot animals underwent surgery for chronic co-implantation of the EC device and CI into the scala tympani. One subject died during surgery due to anesthesia failure. The other four received EC devices containing GDNF (n=3) or BDNF (n=1) producing ARPE-19 cells. In all implanted animals the handling of the EC devices allowed for an unproblematic co-insertion of the device into the scala tympani implanted with a cat CI. The design and size of the EC devices does not have to be further improved for in vivo cat implantation.
In all four implanted animals electrode impedances were measured directly after implantation, showing that the CI was intact. Following implantation, daily impedance measurements (via DibBox and Maestro Software from Med-El) were performed. During the first two weeks all electrodes exhibited high impedances indicating wire breakage and / or contacts being in non-conductive surroundings (i.e. being out of the cochlea). X-ray images were taken, showing that in all cases except the last implanted cat, the electrode had moved out of the cochlea. Based on this finding and due to the fact that the evaluation of all explanted implants showed wire breakage, the CI design was changed from experiment to experiment to eliminate all possible negative influences on wires and electrode placement. The cat CI development was led by Partner 3 (Med-El, Austria). Additionally, the surgical technique has continuously been improved based on the results and in accordance to suggestions from other working groups.
Two different approaches of CI implantation have been assessed: a subcutaneous and a transcutaneous approach. The subcutaneous electrode is similar to the human CI electrode in that all parts are implanted into the subject and coupling to an external stimulator is done via a subcutaneous coil. In the transcutaneous approach, which is the common approach in animal research, the electrode ends in a connector, which is reaching through the skin. Both approaches will be further explored to develop the best CI electrode for long-term chronic electrical stimulation in the cat.
Chronic preclinical tests in deafened cats for determination of long-term safety and functionality (Verena Scheper, Wiebke Konerding, Heike Janssen, Peter Hubka, Andrej Kral, Thomas Lenarz, Jenny Ekberg, Lars Wahlberg, Jens Tornøe)
In vivo testing of combined EC and CI implantation
The in vivo tests aimed to investigate the structural and functional effects of the selected factors released by encapsulated cells and devices produced within WP1 and WP2, using an animal model mimicking a severely deaf patient fitted with a cochlear implant. It is hypothesized that the electrical responsiveness of the auditory system will be significantly higher in animals treated with factor-releasing encapsulated cells.
The aims of the second half of the work package 4 schedule were:
1. To evaluate long-term biocompatibility and neuronal preservation using histological methods
2. Proof of long-term stability and functionality of EC-devices in vivo
All chronically implanted (CI and EC device) animals underwent the following procedures: deafening, training, implantation to the left ear, chronic measurements of CI functionality (impedances and hearing/auditory brainstem response threshold), acute CI implantation of left and right ear, final recording and histology of the cochlea.
The chronic implantation and electrical stimulation has been described in detail in Deliverable 4.3. In short, 21 cats were unilaterally co-implanted with an encapsulated cell (EC) device and a cochlear implant (CI) into the left scala tympani via the round window. The cats were assigned to three groups which received different treatment combinations of EC devices secreting glial cell-line derived neurotrophic factor (GDNF) and CI electrical stimulation (ES). Additionally, normal hearing animals served as controls.
The electrical stimulation of the +GDNF/+ES group (N = 7) started 14 days post-surgery. Cats were stimulated by passive acoustic stimulation (acoustic environment of the animal facility and additional radio broadcast) via the OPUS speech processor (MED-EL) and subcutaneous PULSAR implant (MED-EL) for 4 hours a day, 5 days a week. After confirming the functionality of the implant via electrically evoked auditory brainstem response (eABR) measurements, the level of stimulation was defined based on behavioural hearing thresholds. One animal was finalized already after 2 months, due to health problems unrelated to the cochlear implants. Thus, results are based on 20 animals (+GDNF/+ES: N =6, +GDNF: N = 8, -GDNF: N = 6) implanted for 6 months.
At the end of the 6-month implantation, we performed eABR measurements to assess the functionality of the remaining spiral ganglion neurons (SGN). For analysis (Matlab and R software) artifacts were blanked (1 ms) and signals filtered (300-5000 Hz).
To assess the overall response strength, we calculated root mean square (RMS) values (1-7 ms post stimulus interval). The input-output function (dependency of RMS value on stimulation level) was fitted with a sigmoidal function (formula 1):
y=a/(1+e^(-b(x-c)) )
where, a=upper asymptote (i.e. RMSMAX) [µV], b=slope [µV/dB], c=inflection point [dB]
The fitting procedure was successful in 14 left (implanted) and 18 right (untreated control) ears. Threshold levels (dB att.) were assessed at 10% RMSMAX and dynamic range (dB att.) was calculated from 10-90 % RMSMAX. The slope (µV/dB) was the maximal slope at the inflection point of the curve. Individual measurements were excluded when the fitted inflection point fell outside the stimulated dB-range.
To assess which parameter was best suited to represent differences in auditory nerve function, we first compared the untreated (right) ears to normal hearing controls. There was no significant difference between hearing controls (N=4) and neomycin deafened animals (untreated ear, N=18) for threshold, slope, or dynamic range. However, RMSMAX was significantly higher in controls than in neonatally deafened animals (meanhearing: 4.47 µV, SDhearing: 1.20 µV, meandeafened: 1.30 µV, SDdeafened: 0.57 µV; Mann-Whitney U test: p=0.003).
For the analysis of treatment effects, we calculated the relative percentile difference in RMS value between treated (left) and untreated (right) ear (formula 2): (left-right)/(right)∙100
Positive values indicated higher values in the treated compared to the control ear. The RMSMAX values were similar between ears, in all three groups (Wilcoxon signed rank test: control: p=0.125 +GDNF: p=1.000 +GDNF/+ES: p=0.625). Still, there was a statistical trend for a positive correlation between ES duration (+GDNF/+ES group) and RMSMAX (r=0.900 p=0.083) with longer chronic ES leading to higher RMS magnitudes at the left (treated) compared to the right ear.
The histological evaluation of the auditory nerve (neuronal preservation) as well as the EC device and fibrous tissue formation (biocompatibility) was described in detail in deliverable 4.4. In short, the animals were transcardially perfused under deep anesthesia and the cochleae were harvested and prepared for plastic embedding. To assess SGN survival, 5 midmodiolar slices at 20 µm distance (i.e. every 5th slice) were analyzed. In each, 4 sections of the Rosenthal’s canal were assessed: lower basal, upper basal, lower middle and upper middle. We calculated median SGN densities per mm2, based on the diameter of the Rosenthal’s canal and the number of SGN with clearly identified nucleus (LAS software, Leica microscope DM1000 & camera DFC 320, Leica GmbH). To assess the time course of SGN degeneration after neomycin deafening, SGN densities of the right (untreated) ear were compared to pilot animals (N = 6) and hearing controls (N = 4). The SGN density significantly declined with duration of deafness (Spearman correlation: r=-0.621 p=0.0003). The data was fitted with an exponential decay (GraphPad Prism, r2=0.932) yielding an Y0 (i.e. hearing controls) of 2119 cells/mm2, declining to a plateau of 237 cells/mm2 after about 8 months of neonatal neomycin deafening, with a half-life of 1.27 months.
We calculated relative percentile left-right differences between ears for SGN density (cf. formula 2), to assess the influence of treatment on auditory nerve survival. We separately assessed basal (median of lower and upper basal turn) and medial (median of lower and upper medial turn) parts of the cochlea. Positive values indicated higher SGN densities on the treated compared to the untreated ear. At the basal part of the cochlea was a significant difference between groups (Kruskal-Wallis test: p=0.0015) with +GDNF/+ES having significantly higher relative left-right difference in SGN density than the +GDNF group (Dunn’s Multiple Comparison test). Control animals and –GDNF animals had similar SGN densities at left and right ears, with relative left-right differences being not significantly different from zero (Wilcoxon signed rank test: control: p=0.0875 –GNDF: p=0.094). The +GDNF group showed a significant shift to negative values (p=0.023) indicating a lower SGN density on the left (treated) ear compared to the internal control ear. The +GDNF/+ES group, however, had significantly higher SGN densities at the left than right ear (p=0.031).
To assess the functionality of the EC device we took two approaches: 1) GDNF production assessment ex vivo and 2) histological assessment of the encapsulated cells in situ (for details see deliverable 4.4). Preimplantation concentration measured via enzyme-linked immunosorbent assay (ELISA) revealed mean production levels of 79.9 ng/24h/device (SD: 50.6 ng/24h/device, N=14). In the 6 EC devises remaining in situ for histological analysis (1 device from –GDNF group), we either found no ARPE-19 cells or cell debris (i.e. without cell nuclei; mean score: 0.7 range: 0-1). Of the EC devices producing GDNF, 9 were explanted at the end of the chronic implantation and were retested for neurotrophic factor release. Of these devices one was found to produce GDNF (animal 1421). This production (14.1 ng/device/day) was reduced to about 13% of the pre-implantation level (110.9 ng/device/day). Taken together, we showed directly (histology) or indirectly (ELISA) that ARPE-19 cells in 14 out of 15 EC devices died during a 6-month implantation time in the scala tympani of the cat. The cat with the EC device showing some remaining GDNF production (#1421) was no outlier in its group (+GDNF+ES), both concerning SGN survival and eABR measurements.
All devices left in situ were found to be encapsulated with a fibrous tissue sheath; either loose or dense (score 4 or 5). The mean score for fibrous tissue was 4.7 (SD: 0.4) with similar values throughout the cochlea (mean+SD: before: 4.7+0.5 midmodiolar: 4.8+0.4 after: 4.7+0.4). We concluded that this encapsulation was restricting diffusion of nutrients and was the most probable cause for the observed ARPE-cell death.
To assess potential effects of the GDNF treatment on fibrous tissue growth, we compared fibrous tissue in the scala tympani between left (treated) and right ear. At the right ear, only small amounts of fibrous tissue were found at the lower basal turn (median: 1, range: 0-1) and none was found in the upper basal turn. This was true for all individuals. At the left ear, we found on average higher amounts of fibrosis. For the –GDNF group a median of 1.5 (range: 0-3), for the +GDNF+ES group 1.5 (range: 1-3) and for the +GDNF group a median of 3 (range: 1-4). The left-right difference in fibrosis was not significant from zero for both the –GDNF and +GDNF+ES group (Wilcoxon signed rank test: –GDNF: p=0.149 +GDNF+ES: p=0.095) indicating similar levels of fibrosis at both ears. The distribution of the +GDNF group was significantly shifted to positive values, indicating a higher fibrous tissue growth at the left compared to the right scala tympani.
We propose that chronic ES was reducing the fibrous tissue amount in the scala tympani, allowing for an initial neuroprotective effect via GDNF delivered by the EC device. We conclude that additional anti-inflammatory treatments and/or changes of the EC device membrane would be feasible to reduce fibrosis and thus allow for a stable long-term deliverable of NTFs via EC devices in CI patients. Although the device was not stable over the 6 months implantation time in the deafened cat cochlea, the potent neuroprotective effect in the +GDNF+ES group clearly marks the potential of EC device treatment for human CI patients.
Potential Impact:
POTENTIAL IMPACTS
The NeuEar project aimed at developing genetically modified cells for encapsulation and long-term expression of selected neurotrophin(s) in a medical device implant prototype that can be combined with a cochlear implant. The NeuEar partners have also optimized safety and functional effects in in vitro and in vivo assays and evaluated the long-term safety and efficacy in animal models of a clinically relevant implant. The consortium has therefore successfully reached its final goal to have an implant prototype and functional data in suitable animal models available that support a continuation towards clinical testing after the NeuEar project has been finished.
Innovative approach
The implantable device represents a true potential for clinical application and future commercialization. This project aimed to be innovative but at the same time drew upon the previous successful technology development by the industrial partners (NsGene & MED-EL) and the research and translational experience of the academic partners (KI and MHH).
Several innovative developments have been brought forward by the project:
• First cochlear implantations of Partner 1’s EC device in guinea pig and cat deafness models as well as human temporal bone, expanding the scope of the technology platform.
• First combination of a cochlear implant with an encapsulated cell device releasing neurotrophic factors.
• Development of new cochlear implants for chronic implantation in cats.
• Development of the EPS concept allowing separate implantations of devices parallel to a cochlear implant (both human and animal).
Promotion of SMEs in Europe
The project, led by the SME NsGene, Denmark, integrates an industrial partner and two academic groups working together to reach the delivery of a therapeutic cell-based encapsulation device that can be implemented in conjunction with a cochlear implant. The consortium has demonstrated that it possessed all the expertise and know-how to develop, design, and validate this chronic implant and bring it close to the market.
Impact for the patients and the society
In Europe overall, more than 80 million persons (around 16%) suffer from hearing loss that is severe enough to adversely affect their daily life and only one in nine persons afflicted with hearing loss is still working. Despite common perception, hearing loss doesn’t only affect the elderly. The reality is that the majority (65%) of people affected by hearing loss are younger than age 65. In the States, 1.4 million children (18 or younger) have hearing problems; and it is estimated that 3 in 1000 infants are born with serious to profound hearing loss (survey from Better Hearing Institute).
Economic benefits
Untreated hearing loss costs Europe 213 billion euro per year, according to the international scientific report ”Evaluation of the Social and Economic Costs of Hearing Impairment”. This equals about 473 euro per year for each adult European. Of course, with an increase of the aging population since many decades, it is foreseen that the current prevalence of at least 3 in 10 people over age 60 having hearing loss will worsen. The great medical need has created a large worldwide market for cochlear implants, which was $725 million in 2008 and is expected to grow to $1.59 billion in 2014 (Neuroreports). Even though the current commercially-available implants restore hearing to great extent, their performance remain suboptimal under difficult hearing conditions (for instance in noisy environments). Furthermore, the established and possibly ongoing degeneration of primary auditory neurons (ANs) is a limiting factor in cochlear implant efficacy. The exogenous application of neurotrophic protein factors can prevent these degenerative processes and is expected to allow a degree a functional improvement. By combining a highly innovative device that will permit the long-term delivery of such neurotrophic factors with a cochlear implant, the project has a high socio-economic value for the European ageing society.
MAIN DISSEMINATION ACTIVITIES
NeuEar teams have placed a particular emphasis throughout the grant period to disseminate their work and scientific highlights to the scientific community, the European Commission, patient associations and public at large. You will find here a summary of the different actions taken:
Dissemination to the Scientific Community
Participation to international conferences, poster presentations and lectures
The full list of national and international conference attendance by the NeuEar teams is available on the ECAS portal.
List of publications
The full list of publications is available on the ECAS portal.
Creation of a NeuEar group on LinkedIn
To increase the visibilty of the NeuEar project within the scientific community, Isabelle Weiss (DWC, Partner 5) has created a NeuEar group on LinkedIn. The link to the NeuEar’s LinkedIn group is available on the NeuEar homepage.
Link: http://www.linkedin.com/groups/NeuEar-4843435?gid=4843435&trk=hb_side_g
22 members including 13 non-NeuEar members are following the NeuEar group on LinkedIn.
Publication of articles on the NeuEar project
• Lars Wahlberg and Jens Tornøe (Partner 1) and Isabelle Weiss (Partner 5) have been invited to write up an article on the NeaEar’s project and objectives by Pan European Networks (www.paneuropeannetworks.com). This jointly written article has been published in June 2013 in the issue 07 of PAN’s Science and Technology Journal (pages 210-211).
• Interviews given by Thomas Lenarz and Verena Scheper (Partner 4), and Jens Tornøe (Partner 1) to International Innovation were published in an article entitled ‘Loud and Clear’ on July 2014.
Development of a flyer to advertise the NeuEar’s research project
Partner 5 (DWC Ltd) in collaboration with Partner 1 (NsGene A/S) has rapidly developed a flyer for NeuEar. The flyer is presenting the research objectives of NeuEar, the teams involved and is acknowledging the EC funding. The flyer was developed in March 2013 to be used as a dissemination tool (see actions in the sections below). This flyer has been translated in French to be distributed to all kids who participated to the NeuEar outreach project in Spring 2015 (see section below).
Dissemination to Members of the European Parliament and the European Commision
The NeuEar flyer has been sent early 2013 to 45 Members of European Parliament together with the first NeuEar press release (published in February 2013).
Publication of press releases
A first press release has been published in February 2013. This press release was presenting the NeuEar scientific project and partners to the European Commission and the scientific community. By placing the link on the NeuEar website, it is also available to the public.
This press release was also published on Cordis – Weblink:
http://cordis.europa.eu/project/rcn/104052_en.html
Publication of the NeuEar’s project goals and objectives on the EC portal for Research and Innovation:
http://ec.europa.eu/research/health/medical-research/severe-chronic-diseases/projects/neuear_en.html
A second press release has been published in April 2015 following the annual meeting of the consortium in Innsbruck on March 2015. This press release was published on the NeuEar website and published on Cordis: http://cordis.europa.eu/news/rcn/123975_en.html
Dissemination to Patient Groups and Charities
In order to communicate with the relevant patients associations and charities that support European citizens impacted by hearing loss, Partner 5 (Isabelle Weiss from DWC) has listed 22 patient organisations throughout Europe using contacts already taken by the PIs of NeuEar and a web search. It has been indeed highly important for NeuEar to take and maintain contact with people suffering from hearing loss to inform them about the goals of the project with the development of this new device and to keep them informed on the research and technology progresses.
These 22 patient organisations listed have been contacted for the first time in March 2013 through a letter by Isabelle Weiss (Partner #5) to inform them about the scientific objectives of the consortium and to present them the first press release and the NeuEar flyer. All associations having their website listed on the NeuEar public website have been informed and did agree to it.
In return, some organisations have placed information on the NeuEar project on their own websites. Some examples are given below:
http://www.observatorioplastico.com/detalle_noticia.php?no_id=232511&seccion=envase&id_categoria=914
http://www.ingenieros.es/noticias/ver/proyecto-neuear-desarrollo-de-implantes-para-restaurar-la-funcion-auditiva/3408
These organisations are kept informed on the progress of the NeuEar project on a regular basis.
Dissemination to the public at large
NeuEar has developed early 2015 an ‘Art & Science’ outreach workshop dedicated to kids from 7 to 14 years old. Three workshops were presented in ‘Youth and Culture Centers’ (Maison des Jeunes et de la Culture in french) in Lyon and Bron (France) in April and May 2015 with a total of 45 children.
The goal of this 2-hour workshop was to intrigue and engage children with the wonders of the human ear and today’s technological developments like the cochlear implant. It was also to sensitize them to the different types of deafness and their origins, and to remind them about the importance of protecting their ears. For this workshop, we have developed several types of hands-on sessions together with a powerpoint document full of videos and pictures for a lively and fun presentation.
During the workshop, we have asked the children to locate the tympanum, the ossicles and the cochlea on a drawing presenting the 3 parts of the ear (external, middle and internal ear). Each section was given on a piece of paper to glue on the general drawing (see Picture 1). We have taken the opportunity of this hands-on session to describe the roles of the different sections of the ear and to explain the propagation of the sound through the ear and up to the brain. We have also explained what was the sound and its characteristics (frequency, decibels).
Kids were then asked to place different sources of noise from their everyday life on a decibel ladder; such as a washing machine, a vacuum cleaner, a classroom, a space rocket, a live concert, etc... This exercise triggered many discussions and questions in groups. Following this, we have created a simplified brain in air-dough to locate the brain region treating the sounds. The air-dough used for this hands-on session is a very light type of playdough drying to solid when left outside the plastic bag. This air-dough was colored so the kids could shape each cerebral lobe with a different color. Using this creative workshop, we have approached the notion of interpretation of sound and ear cell degeneration.
At the very end of the workshop, Dr Fabien Seldran from MEDEL, France presented a replica of a Medel cochlear implant to the children and explained the cochlear implant technology. 3-D model of the inner ear and cochlear implant insertion were also presented to the children to support our explanations.
Each child returned home with her/his own drawing of inner ear, her/his created model of brain, the NeuEar brochure translated in French and a flyer to help her/him creating new brain models at home. Both the NeuEar brochure and the workshop flyer have the NeuEar and EU logos. Hopefully these documents have triggered discussions at home with the parents!
EXPLOITATION OF RESULTS
Exploitation plan
NsGene has successfully brought the EC implant technology from research into clinical development for Alzheimer’s disease. In this process, the SME has gained insight and expertise in the necessary regulatory, manufacturing, and product development process. NsGene has therefore the ability to move the project beyond the 3 years of the initial phase. It is the intention of the consortium to apply for Phase Ib clinical testing based on successful completion of this 3-year project. If the EC implant is used in conjunction with a cochlear electrode, it is anticipated that the commercialization will be done in a partnership with MED-EL as this industrial partner has already established a commercially successful product in this therapeutic domain. The EC implant will also form a new product platform for regenerative therapies to the inner ear and as such could be developed into several pertinent products over the next few years.
List of Websites:
Dr Lars Wahlberg, COO, NsGene A/S
Tel: +45 2360 3198
E-mail: luw@nsgene.dk
Sixteen percent of adult Europeans suffer from hearing loss, great enough to adversely affect their daily life. Over the age of 80, 50% of the population is suffering from hearing loss. A large portion of this population is affected by sensorineural hearing loss (SNHL), the consequence of a progressive degeneration of the primary auditory neurons (ANs), the afferent neurons of the cochlea. The only current therapeutic intervention for these patients is the use of a cochlear implant (CI), a neural prosthesis designed to directly electrically stimulate the ANs. Despite improvements in CI technology, the hearing experience with these devices remains far from normal, with many patients reporting difficulties discriminating speech in noise and poor perception of temporally encoded sounds such as music.
During the course of the grant, NeuEar partners have designed a neurotrophic cochlear implant – a novel neural prosthesis - to provide both electric auditory cues and regenerative neurotrophic factor(s) to severe-profoundly deaf patients. As the ongoing degeneration of ANs that occurs over time is a limiting factor in current cochlear implant efficacy, the ANs have been the target cells in the implant design strategy. The project aims were to develop an encapsulated cell (EC) therapy device capable of long-term intracochlear delivery of neurotrophic factors to prevent the degeneration of ANs, and to further combine this new EC device with a cochlear electrode implant for joint implantation in the cochlea.
As a first step, several clones based on a human retinal pigment epithelial cell line were developed to express high levels of either Glial cell-line Derived Neurotrophic Factor (GDNF) or Brain-Derived Neurotrophic Factor (BDNF). To test the clones in a relevant animal deafness model, a novel encapsulated cell device was engineered to allow for cochlear implantation in guinea pig and cat. In a one month (short term) guinea pig study, we showed that the encapsulated cells expressing either GDNF or BDNF successfully protected the animals from normally deafening, injected doses of neomycin. These very encouraging proof-of-concept results show that intracochlear delivery of neurotrophic factor from an encapsulated cell device can indeed provide a positive effect to diminish hearing loss.
In a following long-term, 6-month cat study, encapsulated cell devices were implanted together with custom-made cochlear implants designed by the consortium. Good results on safety and the lack of adverse effects in the cats were obtained, but the cochlear implants were generally broken over the six months due to the natural movements of the cats. Furthermore, encapsulated cell devices implanted over the same time were found to be rejected by the cats, leading to their loss of function. Compared to the good results obtained in guinea pigs, this indicates that the cat deafness model could be difficult to work with for chronic cochlear implantations.
In parallel to the in vivo studies, several design concepts for a final clinical product were made, resulting in functional prototypes that can be employed in coming clinical studies. These designs are currently being patented and will, based on the very encouraging NeuEar data, be used in a clinical Phase I trial planned to start shortly after the end of the NeuEar project.
Project Context and Objectives:
Project context
Hearing loss (HL) is the most prevalent sensorineural disorder in developed countries and the most common birth defect. One of every 500 newborns is affected with bilateral congenital sensorineural hearing loss (SNHL) ≥40dBHL. This number rises to 2.7 per 1000 before the age of 5 and 3.5 per 1000 during adolescence (Morton, 1991). The SNHL typically occurs following damage and loss of hair cells - the sensory cells in the cochlea of the inner ear - which, in response to sound, convert the mechanical vibrations into nerve impulses in the primary auditory neurons (ANs). Widespread hair cell loss results in severe-to-profound SNHL (Nadol et al., 1997). The only therapeutic intervention for these patients is the use of a cochlear implant (CI), a neural prosthesis designed to directly electrically stimulate the ANs. The CI system consists of an external unit and an implanted, internal unit. The behind-the-ear external processor with ear hook (1) and a battery case (2) uses a microphone to pick up sound, converts the sound into a digital signal, processes and encodes the digital signal into a radio frequency (RF) signal, and send it to the antenna inside a headpiece (3). The headpiece is held in place by a magnet attracted to an internal receiver (4) placed under the skin behind the ear. A hermetically sealed stimulator (5) contains active electronic circuits that derive power from the RF signal, decode the signal, convert it into electric currents, and send them along wires (6) to electrodes (7). The electrodes are assembled on the silicone rod (electrode array) and inserted into the cochlea, where they stimulate the auditory nerve (8). The nerve conveys the signal to the central nervous system, where the electrical impulses are interpreted as sound.
Despite improvements in CI technology, the hearing experience with these devices remains far from normal, with many patients reporting difficulties discriminating speech in noise, and poor perception of temporally encoded sounds such as music. Part of the inefficiency of cochlear implants is related to a poor interface with and progressive degeneration of the ANs, ultimately leading to significant neuronal loss after long periods of deafness. In addition, experimental animal studies indicate that on-going AN degeneration compromises the efficacy of the cochlear implant.
The loss of endogenous neurotrophic factors normally expressed by hair cells, such as brain derived neurotrophic factor (BDNF) and neurotrophin-3 (NT3), initiates this AN degeneration. Numerous studies have demonstrated that intracochlear infusion of these neurotrophins can support AN survival in animal models of deafness. Furthermore, when combined with chronic electrical stimulation via a cochlear electrode, exogenous neurotrophin treatment results in significantly enhanced AN survival compared to neurotrophin treatment alone (Shepherd et al. 2005). The novel EC implant to be developed in this project takes advantage of cell-based de novo synthesis of the therapeutic neurotrophin and allows for the chronic production of bioactive protein. Hence, the encapsulated cell device aims to combine the advantages of gene therapy with that of the safety of a removable device, attached to the cochlear electrode itself or as a separate implant.
The EC technology consists of an elongated device that contains a genetically modified human cell line enclosed behind a semi-permeable hollow fibre membrane that allows for the influx of nutrients and the outflow of internally produced therapeutic factor(s), but does not allow for the direct contact of the cell line and the host tissue.
The implant provides for long-term factor secretion of de novo synthesized protein while allowing for its retrieval. In addition, the permselective membrane prevents immune rejection of the allogeneic cell line, allowing for the same parental cell line bank to be used to produce the various therapeutic cell lines in all products and individuals. The long-term stability (1 year) of this delivery system in the brain has been demonstrated with an EC therapy product secreting nerve growth factor (NGF). This product is currently in clinical trials for the treatment of the cholinergic neuronal degeneration in patients with Alzheimer's disease. A total of 10 patients have been implanted with up to four NsGene clinical devices per patient and the technology has thus been through technical, manufacturing, safety, functional, and regulatory proof of concept. With the proposed research and development with the collaborators and work packages in this project, it is anticipated that the NsGene encapsulated cell technology can be modified and developed for use in the inner ear within the proposed three-year timeframe.
Project objectives
The NeuEar project aimed to develop genetically modified cells for encapsulation and long-term expression of selected neurotrophin(s) in a medical device implant prototype that can be combined with a cochlear implant. The NeuEar partners have also evaluated and optimized safety and functional effects in in vitro and in vivo assays and evaluated the long-term safety and efficacy in a guinea pig and cat animal models.
The final goal was to have an implant prototype and functional data in suitable animal models available that support a continuation towards clinical testing after the NeuEar project has been finished in the proposed 3-year period.
The specific objectives of the NeuEar project were to:
(i) develop genetically modified cells for encapsulation and long-term (>6 months) over-expression of selected neurotrophin(s) in vivo
(ii) develop EC implant prototypes that can be combined with a cochlear implant
(iii) evaluate and optimize safety and functional effects in in vitro and in vivo assays
(iv) evaluate the long-term safety and efficacy in a large animal model of a clinically relevant implant
(v) implement an efficient dissemination structure, including a patent strategy to enable partnering and fund-raising for further clinical development, regulatory approval, commercialization, and marketing.
References
• Morton NE. Genetic epidemiology of hearing impairment. Ann NY Acad Sci 1991;630:16-31
• Nadol JB Jr. 1997. Patterns of neural degeneration in the human cochlea and auditory nerve: implications for cochlear implantation. Otolaryngol Head Neck Surg. 117:220-8.
• Shepherd RK, Coco A, Epp SB, Crook JM. (2005) Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J Comp Neurol. 486(2):145-58.
• Zeng, F.; Rebscher, S.; Harrison, W.; Sun, X. & Feng, H. (2008) Cochlear Implants:System Design, Integration and Evaluation. IEEE Rev Biomed Eng, 1, 115-142.
Project Results:
In the first half of the NeuEar project, a particular emphasis was given for the selection of the most appropriate cell lines and neurotrophic factors to be used in the EC device. In this last half of the project, characterization of the developed cell lines was further continued by Encapsulated Cell (EC) device implantation in both guinea pig and cat animal deafness models to identify a candidate neurotrophic factor and cell line for treating sensorineural hearing loss. First results came from guinea pigs where both cell lines tested – one secreting GDNF and one BDNF – were shown to protect spiral ganglion neurons from the deleterious effects of neomycin injection over the course of one month. Compared to untransfected ARPE-19 cells (not expressing exogenous neurotrophic factors), the GDNF #125 cell line showed a significantly smaller increase in eABR threshold with a very narrow standard deviation among samples. The BDNF #202 cell line also caused eABR thresholds to remain low, but variation in eABR thresholds was larger than for the GDNF line. Based on these very encouraging data, the GDNF #125 cell line was chosen as candidate cell line for the remainder of the NeuEar project with BDNF #202 as backup cell line. The CDNF #605 line that was a contender during the first project period was finally abandoned, as several attempts of redesigning the amino acid sequence to increase expression ultimately failed.
As result of the short-term guinea pig data, the long-term, chronic cat study was designed to focus on GDNF #125, with a few pilot cats implanted with the BDNF alternative. Where the short term guinea pig study made use of already well-established cochlear implant electrodes, CIs for the cat study had to be custom made by Partner 3 in collaboration with Partner 4. It quickly became evident that designing a durable device able to maintain its function over six months in the very agile cat was to be a problem. Most electrodes failed shortly after implantation and it took several design iterations and implantation procedure optimizations to increase CI longevity sufficiently, which in turn decreased the number of functioning animals available for analysis at the study end. Most encapsulated cell devices filled with GDNF #125 cells unfortunately did not show any neurotrophic factor expression after 6 months in vivo. The reason for this is most likely a strong fibrotic reaction to the implanted devices, as could be observed from the histologic analysis of explanted cochleae. Combined with the electrode challenges, we were therefore not able to obtain as robust data in the cat as in the guinea pig. Nonetheless, SGN protection was still evident in the experimental group combining GDNF device implantation with electrical CI stimulation (ES). GDNF devices without ES did not show a protective effect, raising the concern that GDNF treatment in itself did not result in a protective effect. However, the number of animals in each group was too low and the electrode variability too high to make a firm conclusion.
In addition to the cat CI design, concept EC and CI devices for human implantation were successfully designed and developed. Two approaches were chosen, one where the EC and CI were fused to each other allowing for a single-implantation procedure, and one approach where the CI and EC were implanted separately. Several iterations of both concepts were tested in cochlear plastic models as well as in human temporal bones. While both concepts were feasible, the advantages associated with the separated approach, in particular with respect to flexibility and little need for CI modification, were found to outweigh the surgically simpler fusion device. For use with the separate EC implantation, an implantation port concept named EPS (Extension of Perilymphatic Space) was developed. Several EPS design iterations were successfully tested in human temporal bone.
Details of the major achievements in each scientific work package are summarised below.
WORK PACKAGE 1 – SELECTION OF NEUROTROPHIC FACTORS AND GENERATION OF CELL LINES
Cell line development (Wahlberg, Tornøe, Ulfendahl, Lenarz, Scheper)
The first task of the NeuEar Work Package 1 concerned the development of cell lines to be used in Partner 1’s (NsGene, Denmark) EC Biodelivery technology for cochlear application. The project started out with a plan focusing on developing ARPE-19 based cell lines expressing the neurotrophic factors brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and Cometin, the latter being a neurotrophic factor discovered recently by Partner 1. NT-3 and Cometin was later replaced with glial cell-line derived neurotrophic factor (GDNF) and cerebral dopamine neurotrophic factor (CDNF). GDNF was chosen for its well-documented trophic effect on a wide range of neurons, including auditory neurons, and the novel factor CDNF was selected as a scientifically interesting and exploratory molecule to test due to its possible effect on auditory neurons combined with very few reports on the application of CDNF for hearing indications thus far.
To be able to generate high-producing cell lines expressing the three neurotrophic factors of choice, plasmid vectors based on Partner 1’s proprietary Sleeping Beauty (SB) transposase expression system were generated. Briefly, the SB system allows for high-level genomic integration in the recipient cells mediated by a co-transfected transposase, leading to strong and stable expression of the introduced transgenes. The completed vectors were tested transiently for their ability to mediate factor expression from ARPE-19 cells, followed by generation of stable cell clones by antibiotic selection. All plasmid preparations used for the cell clone generation was made under GMP at Plasmid Factory, Germany to ensure that no adventitious agents were introduced.
The stable transfections resulted in a large set of clones to be characterized for their neurotrophic factor production levels. While expression of GDNF and BDNF was high (up to 12 µg factor / 106 cells / 24 h), CDNF expression from isolated clones was about 50 times lower, on average. CDNF transfection was repeated several times with similar results. As the CDNF levels obtained from the ARPE-19 clones could not reach physiologically relevant levels, the project focus was directed to GDNF and BDNF clones. Here, GDNF clone #125 and BDNF clone #202 were chosen as candidate clones for further studies.
Cell line selection (Wahlberg, Tornøe, Ulfendahl, Lenarz, Scheper)
After having identified the high-producing clones GDNF #125 and BDNF #202, further characterization was needed in order to qualify the clones for the later in vivo experiments in deafened guinea pig and cat. From the ELISA data we knew that high levels of immunogenic protein was released from the cells, but further testing of the biological activity was warranted. In theory, truncated protein variants expressed from the ARPE-19 clones could be ELISA positive while having lost their biological effects.
Furthermore, the in vitro factor release level is but one of several important properties of a clinically relevant cell line. Equally important, a suitable cell line must show good growth properties and be able to survive in the more restrictive in vivo environment after implantation. This was tested by striatal implantations in Göttingen minipigs and rats. Both clones showed sustained factor production after up to three months implantation of the EC device. Immunohistochemistry of rat striatum implanted with GDNF #125 showed that GDNF produced from the devices did readily diffuse into the target tissue. Furthermore, the biological activity of GDNF was assessed by staining for tyrosine hydroxylase (TH) which is commonly known to be upregulated by GDNF. An increased TH staining level in the implanted side compared to the control side confirmed that the GDNF secreted from the implanted EC device was indeed active.
Due to technical issues at Partner 2, the in vitro testing of BDNF #202 was delayed. To carry on testing the clone anyway, it was included in a guinea pig evaluating the AN-protective effects of the encapsulated cells. Here, both the GDNF and BDNF clone showed good protective effects compared to control cell not expressing measurable amounts of neurotrophin.
In conclusion, both the GDNF and BDNF clones produce such high levels of neurotrophin that therapeutically relevant release levels can be achieved. For the CDNF clones, however, we have not been able to reach this goal. Despite a large number of attempts, the best of the isolated CDNF clones have expressed almost two orders of magnitude less factor than their GDNF and BDNF counterparts. Consequently, we have moving forward with GDNF and BDNF without CDNF. Solving the CDNF expression issue calls for redesign of the entire expression system to accommodate this factor. Due to the high relevance and novelty of CDNF application for deafness, Partner 1 and 3 continued to pursue this goal and finally showed that therapeutically relevant doses of CDNF were indeed possible to achieve in the in vitro SGN assay.
WORK PACKAGE 2 – DEVICE DESIGN AND PROTOTYPING
Design and manufacturing procedures for animal and clinical devices (Wahlberg, Tornøe, Ulfendahl, Jolly, Mistrik)
All materials employed for device construction are made from clinical grade materials with excellent biocompatibility to minimize tissue reactions after implantation. EC biodelivery devices are made from hollow fiber polysulfone membrane. The membrane dimensions are tailor-made to accommodate the application. The brain device currently employed by Partner 1 (NsGene, Denmark) is made from a membrane developed by NsGene, having an outer diameter of 700 μm and an inner diameter of 550 μm. The molecular weight cut-off is 280 kDa. This membrane configuration will be used for developing the human clinical prototypes in the NeuEar project.
For guinea pig and cat cochlear application, Partner 1 has developed a smaller-diameter polysulfone membrane with a similar molecular weight cut-off. The outer diameter is 400 μm and the inner diameter is 300 μm, resulting in a thinner wall with better diffusion properties. The length of the guinea pig device is 3,5 mm long. For cat cochlear application, a 4 mm long device with the same diameter is used. As scaffold material, a multi-filament yarn material from Swicofil, Switzerland is used, providing the ARPE-19 cells with an extensive surface area on which to attach. Both ends of the device are sealed by medical grade, UV-polymerizable glue.
Cell-containing devices are maintained in serum free medium at 37°C, 5% CO2 until ready for implantation. Manufacturing, filling and sterilisation of EC devices is done entirely independently from that of the cochlear implants to provide a simple functional assembly of the two devices before surgery.
The cochlear implant for the clinical device is the standard, commercially available, CI system manufactured by Partner 3 (Med-El, Austria). Two design approaches for the clinical device have been decided. In the first approach, EC and CI devices are implanted independently. The CI electrode array is implanted first, followed by the EC device. In the second scenario, the EC device is assembled on a modified electrode array of the CI system by a surgeon prior to implantation using a simple clipping mechanism, and both devices are implanted simultaneously. This design is still in the prototype phase.
Neurotrophic cochlear implant for guinea pig implantation (Wahlberg, Tornøe, Jolly, Mistrik)
Guinea pig specifications
The straight part of the guinea pig cochlea in its basal turn is about 6 mm long. Theoretically, it can accommodate a device with a length of up to 5 mm with an outer diameter of 0.6 mm. The electrodes for electrical stimulation are made from 90% platinum and 10 % iridium and are coated with Teflon for insulation. The diameter of the stimulus electrode is 0.075 mm and for the ground electrode the diameter is 0.125 mm. For this particular use in the combination with the EC device, a small ball (0.2-0.3 mm) is burned on the stimulus electrode by using oxygen gas to get a bigger surface when stimulating the auditory nerve. The stimulus electrode is inserted 2-2.5 mm through the round window into scala tympani, the other end of the electrode is stripped from Teflon and soldered to a percutaneous connector. The ground electrode is stripped from insulation approximately 20 mm at both ends, one end is placed along the wall in the middle ear cavity, and the other end is soldered to the same percutaneous connector as the stimulus electrode.
Device construction
A biodelivery device suitable for implantation in the guinea pig cochlea was constructed using the 400 μm membrane developed specifically for NeuEar project. The device is 3,5 mm long and can be placed in the guinea pig cochlea base without disrupting the organ tissue. An inert device handle has been developed to facilitate positioning in the cochlea. The active part of the device is filled with the neurotrophin-producing cell line to be ready for implantation.
Neurotrophic cochlear implant for cat implantation (Wahlberg, Tornøe, Jolly, Mistrik)
Cat specifications
The straight part of the cat cochlea in its basal turn is about 6 mm long and the diameter is about 2 mm. It can accommodate two 5 mm long devices with outer diameter of 0.6 mm each, allowing for parallel implantations of the EC device and CI electrode array. The EC device cannot be bent due to the rigid nature of the polysulfone membrane used for device manufacturing, whereas the electrode array of the CI is made from silicone elastomer providing high flexibility. The EC device is 4 mm long and has an outer diameter of 400 μm. The electrode array for the CI is 8 mm long and the diameter is 0.3 mm at the tip and 0.6 mm at the end.
Device construction
The EC device with cat specification is constructed using manufacturing procedures described above. The CI electrode array was made from platinum and iridium wires coated with Teflon. Each wire is connected to a single contact in the array tip and embedded in silicone using a standard animal-grade aluminum mold. The diameter of the array is 0.3 mm at the tip and 0.6 mm at the end. The length of the array is 8 mm. The wires are 20 mm long. Each of them is connected to a 9-strand platinum wire, which is 150 mm long.
The device was manufactured in two variants. In the first design, the multi-stranded wires are connected to the pins on a transcutaneous connector. The Pulsar stimulator is attached to the second part of the connector. This variant is intended for the implantation of the electrode array, while the stimulator being carried out externally by animal in a backpack (transcutaneous approach). In the second design, the connector is omitted and the multi-strand wires are connected directly to the Pulsar stimulator. This design is intended for subcutaneous implantation of the entire electrode system, including the stimulator.
In the current EC device / CI design, the two devices are kept separate and the EC device is not assembled on the cat electrode array. This design has been adopted after initial tests in temporal bones from cat cadavers, where sequential implantation of the two devices proved to be more efficient. This corresponds to the first of the two clinical approaches. First, the electrode array is inserted through the opening in the round window, after which the EC device is introduced through the same opening in parallel to the electrode array.
Clinical prototypes of neurotrophic cochlear implant
Combined device design
For the construction of the prototype of the neurotropic implant a novel design of the electrode array was needed to be developed to accommodate also the EC device inside the scala tympani (ST) with a limited volume. This design carried a reduced number of electrode contacts, 12 instead of 24, which are organised in one line. In this design, each contact corresponds to a single electrode, and the surface is increased to 0.14 mm2. The planned diameter of this electrode array is 0.6 mm and 0.4 mm at the cochleostomy and tip, respectively. The reduced diameter from 1.3 mm to 0.6 mm is to allow a sufficient space in the basal region for the attachment of the EC device. The EC device was planned to be attached to the array at its basal end at close proximity to the cochleostomy, by a clipping mechanism. For biocompatibility reasons, such clips were planned to be made from silicone, the same material, from which the rest of the array was also made.
The planned clipping mechanism made from biocompatible silicone was also attached to the electrode array and its functionality tested. The extensive testing revealed that this type of the attachment is not optimal, because it increases significantly the diameter of the array in the area where the EC device is attached. This would be detrimental for the insertion, which must be maximally atraumatic. To overcome this problem, an alternative design was invented, based on the stent concept. The stent-like half-cylinder is integrated into the silicone rod to accommodate the EC device inside the electrode array without increasing significantly the diameter.
The first prototype of the scaffolding was made from a 50μm thick platinum sheet with laser-cut openings in it. The sheet was bent to semi-circular form and incorporated into the mould for electrode fabrication. The wires running to electrodes positioned in more apical electrodes are also visible. The implant was extensively tested using the plastic model of the human cochlea. Unfortunately, the repeated insertions revealed that the platinum-sheet based cage is too rigid for atraumatic insertion.
Therefore, the platinum-based half-cylinder was replaced with a polymer mesh tube made from polyimide. The final configuration is with the polymer cage assembled on the electrode array. The EC device is partly exposed. The cage was made in several versions of different length: 7 mm to accommodate the 5-mm long EC device and 5 mm to accommodate the 4-mm long EC device.
Mechanical and insertion tests using the plastic model of human cochlea and artificial temporal bone revealed that also in this configuration the device is relatively rigid and the insertion into a real temporal bone might not be sufficiently smooth to prevent any damage to cochlear structures required for the residual hearing preservation. Therefore, in the final prototyping step the cage was replaced by a more flexible cage, made from the same material, but with the central part cut out to reduce rigidity.
The implantability of both prototypes was tested in cadaveric temporal bones by partner 4. Full insertion was achieved with both prototypes, but only after the electrode array segment with the first 4 electrodes was cut away. Consequently, the maximal possible length of the electrode array was determined to be 24 mm, and not 28 mm of the original design. Therefore, the electrode array was shortened to 16 mm, to provide a sufficient margin for any variation in the cochlear length, and electrodes were redistributed accordingly along the length of the array.
Furthermore, the histological analysis of the cochlea implanted with any of the two designs revealed the full-cage design as less traumatic, even if also in this case some bending of the basilar membrane was observed. In summary, the testing indicated that using the shorter electrode array, Flex16 design, with integrated the full polymeric cage, would be the best design for the combined device, in which the EC device and CI electrode array are integrated into a single implantable device.
Finally, the second prototype of the combined device design was manufactured with a polymer cage of reduced size to accommodate the smaller EC device used in animal experiments in WP3 and WP4 (diameter 0.4 mm, length 4 mm). In principle, such EC device could be also clinically relevant, if the therapeutically agents produced in reduced quantity still have protective/regenerative effect on the target tissue. The dimensions of the cage are: diameter 0.6 mm and length 7,3 mm. The overall diameter of the implant is 1.1 mm, which is smaller than 1.3 mm for the Standard electrode array design (at the level of the cochleostomy entry) and the implantation of this version should be very atraumatic.
Separated device design
The prototyping of the second design, with the EC device and CI electrode array implanted into the cochlea separately, was based on the feasibility testing in cadaveric temporal bone carried out in the first reporting period. After demonstrating the feasibility of parallel independent implantation of both the EC device and CI electrode array into the scala tympani of the cochlea (or of the EC device alone), the question of the independent withdrawal of the EC device, while keeping the CI electrode array in place, was investigated. For that reason, a harbouring port, made from titanium for biocompatibility, and permanently fixed in the temporal bone opening drilled by a surgeon next to the round window was developed.
The thickness of the promontory region of the temporal bone in the target area was measured from the CT scan of an implanted cochlea and estimated to be between 1.6 and 1.7 mm. Therefore it is anticipated that 2 mm long opening must be drilled through the bone to reach the scala tympani. The port was therefore designed accordingly. The specifications (2.2 mm length, 1.3 mm diameter) and the first prototype made from titanium. The port has an extra handle for correct positioning of the sliding part in the bone. The insertion of the port was first tested in a plastic model of human cochlea. It demonstrates that there is enough space to accommodate both devices. Next, the port was tested in the temporal bone. The insertion was done by partner 4. The cochleostomy for the port was drilled next to the round window, the port was fixed in it and the EC device (Length 7 mm) inserted. The Flex16 electrode array was inserted independently through the round window. The consequent histological analysis of implanted temporal bone revealed that 1) the cochleostomy needed for a port of these dimensions was too large (too traumatic) and 2) that the handle was not needed for the correct orientation of the port in the bone. On the contrary, it made the insertion too cumbersome. Instead the screw-thread should be added for better fixation in bone.
These observations were taken into account in the design of the second prototype of the port. The inner diameter was reduced (1.17 mm), the lower part, fixed in the bone, made longer (1.45 mm) and a screw-thread added. The temporal bone insertion test revealed that the dimensions are still unnecessarily large and that the helical configuration of the screw-thread is needed for firmer fixation.
The final prototype (version III) with the helical screw-thread and even smaller inner diameter (0.8 mm). It bears an incision in its upper surface for mounting the port into the bone using a standard microsurgical screwdriver. A mark (2 round spots) is also included to facilitate the correct orientation.
The EC device bears in this final configuration a concept handle made from PEEK for easier handling and hermetical sealing of the port. Such handle could be used also for the device retrieval, if needed. In addition to the port design for the EC device with 0.7 mm diameter, the second version for the smaller EC device (diameter 0.4 mm), used in animal experiments, was also produced to demonstrate the feasibility of the manufacturing, if such smaller device would be used clinically.
Device design
During the second half of the project partners 1 and 3 have closely collaborated on the development and testing of two versions of the clinical neurotrophic implant. In the first configuration the EC device and the CI electrode array were integrated together to form a single implantable device (Combined device design). In the second configuration the designs were modified to facilitate independent but parallel implantation of the two devices into the cochlea (Separated device design). The feasibility of implantation was tested in cadaveric temporal bones by partner 4.
In its final design, the first configuration features the polymeric mesh permanently attached to the CI electrode array of reduced diameter (0.6 mm) and length (16 mm) to accommodate the EC device. Two versions were made to fit a possibly different diameter of EC device (0.7 and 0.4 mm). For the second configuration, the port permanently fixed in the promontory region of the temporal bone was designed, to facilitate independent insertion and retrieval, if necessary, of the EC device through cochleostomy. The CI electrode array is inserted through the round window in this case. Again, also this configuration was prototyped in two versions to accommodate two possible diameters of the EC device.
WORK PACKAGE 3 – FUNCTIONAL TESTING IN VITRO AND IN VIVO
Cellular effects of factors released by encapsulated cells on inner ear tissue (Wahlberg, Tornøe, Ulfendahl, Dash-Wagh, Jolly, Mistrik, Lenarz, Scheper, Hoffmeister, Kristof)
The objective of this task is to demonstrate that factors released by selected clones give positive neuronal effects. This has been addressed by testing the survival of auditory cells exposed to factors produced by cell clones used in encapsulated cell implants. The experiments were performed using spiral ganglion cell cultures to analyse the neuroprotective effect of ARPE-19/ NT cells producing BDNF, GDNF or CDNF compared to control and conditioned medium derived from ARPE-19 cells.
The effects of BDNF, GDNF, and CDNF secreted by the modified ARPE-19 cells on the survival and growth of auditory neurons was investigated using organotypic cultures. The in vitro cultures from the spiral ganglion provide a system that allows for a faster and easier screening of the protective effects of the different neurotrophic factors (NTs). These cultures are physiologically closer to in vivo than the dissociated cells. The effects of NTs will be evaluated in terms of the number surviving neurons, and neurite outgrowth and lengths using immunocytochemistry, microscopy and image analysis.
Survival and growth of auditory spiral ganglion (SG) neurons exposed to factors released by encapsulated cell (EC) devices in co-cultures
Using in vitro spiral ganglion explant cultures were used to analyse the neuroprotective effects of ARPE-19/NT cells producing BDNF, GDNF or CDNF. The release of NTs from individual devices was tested using enzyme-linked immunosorbent assay (ELISA) prior to experiments and in the presence of auditory neurons. The levels of CDNF release were very low compared to the production of BDNF and GDNF devices. The ARPE-19/GDNF and CDNF devices showed less variation in NT production compared to BDNF devices. Explants of the spiral ganglion were co-cultured in 4-well chambers together with NT-producing EC devices, placed at a distance from the tissue of about 2.5-3 mm. After 4 days of incubation, the cultures were fixed and immunolabeled with antibodies against neurofilament and tyrosine kinase receptor B (TrkB). Using fluorescence microscopy the size of explants was measured and the number of neurofilament 200 and TrkB positive neurons was estimated as well as the number and length of the neurites.
Consistent with previous studies, co-culture of neonatal SG explants with BDNF-producing EC devices resulted in a significant increase in the number of neurons in SG explants compare to control, CDNF and GDNF. In addition, the number of TrkB positive neurons were also higher compared to other conditions. To our knowledge, this is the first report of a neuro-protecive effect of CDNF on the SG neurons in vitro. CDNF significantly enhanced the survival of SG neurons (both NF200 and TrkB positive) as compare to control and GDNF. Although low concentrations of CDNF were produced by the EC devices, the effects were similar to that of BDNF devices.
Co-culture with GDNF-producing EC devices slightly but not significantly increased the neuronal survival in the SG explants, which is in contrast to previously published in vivo reports. The difference could be attributed to the different model systems. However, our data is in line with effects reported by other groups on dissociated neonatal or adult SG neurons where GDNF showed no significant neuro-protective effect.
The neuritogenic effects of the released NTs were analyzed by quantifying the number of neurites extending from each explant. The untreated explants showed the fewest neurites while explants co-cultured with BDNF producing EC devices exhibited the highest number of neurites. CDNF also significantly increased the number of neurites extending from each explant. GDNF-releasing EC devices, despite producing amounts of GDNF (> 100 ng/ ml/ 24 hr) expected to have a neuritogenic effect, did not promote neurite outgrowth in the observed SG explants. The length of neurites was strongly increased in presence of BDNF- and CDNF-producing EC devices while GDNF showed a marginal and non-significant increase.
In conclusion, the results indicate that the NT-EC devices produce similar amounts of NT that have previously been reported to induce survival and neurite outgrowth from the SG neuron both in vivo and in vitro. BDNF and CDNF producing EC devices displayed strong neuroprotective and neurogenic activity on SG neurons in the explant cultures. In contrast, GDNF EC devices showed no influence on either survival or neurite outgrowth of SG neurons. A novel finding of this study is the influence of CDNF on SG neurons.
Further studies on CDNF to follow up novel findings in Task 1
As reported above, it was observed that the CDNF released from the encapsulated cell device (ECD) promoted neuronal survival and neurite outgrowth in the spiral ganglion explant. Interestingly, despite the low concentration of CDNF released by the ECD (compared to BDNF released by ECD), the strong neurogenic effects suggested a higher efficiency of CDNF. To induce these effects in spiral ganglion, the neurons should express the receptor of CDNF. Unfortunately, there are no reports about the expression of CDNF in inner ear tissue, and the receptor for CDNF is yet unknown. We performed PCR and immunohistochemical studies to investigate the expression of mRNA and protein CDNF in the postnatal rat cochlea and spiral ganglion explants. Real-time PCR results showed that the transcript of CDNF is expressed in the cochlea during early development. However, expression in the adult is higher compared to the early postnatal ages. The immunohistological studies reveal the expression of CDNF protein in the spiral ganglion neurons particularly in the neurites; both in the P2 cochlea and in the explant cultures prepared from the P5 rats. There was no or very marginal expression of CDNF in the cell bodies of the spiral ganglion neurons in the explants. Similarly, the hair cells exhibited no CDNF expression in the organ of Corti. High expression was observed in the neuronal terminals and in the postsynaptic zone under the hair cells. This expression patterns suggest that spiral ganglion neurons express the CDNF and possibly its receptor, thus, can respond to the exogenous CDNF.
Effects of electrical stimulation on factor-releasing encapsulated cells and spiral ganglion neurons in vitro (Wahlberg, Tornøe, Ulfendahl, Dash-Wagh, Jolly, Mistrik, Lenarz, Scheper, Hoffmeister, Kristof)
During the first project month we tried to establish the electrical stimulation of ECs using the Zahner System which is, for single electrical stimulation of seeded cells, established in our lab. Unfortunately we failed and needed to establish an alternative method for EC stimulation. With help from partner 2, MED-EL, who provided a DIB-Box we were able to build up a new stimulation system which now allows us to electrically stimulate 4 wells in parallel using the same stimulation parameters.
In December 2013 the first electrodes with a “clip-on” system for ECs build from MED-EL arrived in our lab and we were able to perform first stimulation experiments using ECs with encapsulated BDNF and GDNF producing ARPE cells. Briefly, ECs were clipped on the electrode, electrode and EC-device were placed in a chamber slide and fully covered with medium. An initial impedance measurement was performed. Afterwards the electrical stimulation (1000pps, 1200µA) started for 12 hours. Every four hours the supernatant of stimulated ECs was collected, frozen and after finalizing the experiment all samples were sent to NsGene for evaluation of neurotrophic factor concentration by ELISA.
The highest factor production was observed in GDNF-producing ARPE-19 cells. Electrode impedance measurements were performed on each time point before medium extraction and no stimulation related changes in impedances where observed for any of the tested cell lines at any of the measurement time points. Based on factor production, these results indicate that the GDNF-producing clone is the most promising candidate.
The effects of surgically implanting an encapsulated cell device into the inner ear in an in vivo animal model (experimentally deafened guinea pigs) (Malte Hoffmeister, Alexander Kristof, Verena Scheper, Wiebke Konerding, Jens Tornøe, Jenny Ekberg, Briannan Bintz, Chris Thanos, Pavel Mistrik, Anandhan Dhanasingh)
To explore possible functional benefits, a stimulus electrode is implanted at the same time in order to electrically evoke an electrophysiological response from the remaining auditory neurons (spiral ganglion neurons) during several weeks following deafness and implantation.
Guinea pigs with normal hearing were experimentally deafened by transtympanic injections of the ototoxic drug neomycin. After 3 weeks, when significant hearing loss was demonstrated by recording acoustically-evoked auditory brainstem responses, the animals were implanted with one of four different types of EC devices (length: 4mm, diameter: 0.4mm):
1. ARPE-19/GDNF
2. ARPE-19/BDNF
3. Untransfected ARPE-19
4. Empty devices (not containing cells)
The devices were placed into the scala tympani in the basal turn of the cochlea. At the same time, an electrode mimicking a cochlear implant (CI) was also implanted through the round window into the scala tympani. The electrode device was only used for eliciting electrically-evoked auditory brainstem responses (eABRs) not for electrical stimulation. To determine changes in SGN electrical responsiveness, eABRs were measured weekly for 4 weeks. After the last eABR measurement, EC devices were retrieved and neurotrophic factor release levels were measured. The animals were sacrificed and the cochleae collected.
After decalcification, the cochleae were prepared for structural analysis in order to estimate the density of remaining SGN. Also the mean sectional area of the SGN was measured.
Electrically evoked auditory brainstem response (eABR)
To investigate if EC device delivery of neurotrophic factors improved the survival of the SGNs, electrical responsiveness was measured weekly during the 4 weeks that the treatment lasted. The eABR results displayed a significant difference between the two groups treated with neurotrophic factors compared to the two negative control groups. Between the two treated groups, the ARPE-19/GDNF group shows significantly lower thresholds than the ARPE-19/BDNF group. There is a considerable variation in the eABR results within the ARPE-19/BDNF group compared to the ARPE-19/GDNF group. It is possible that this due to the fact that the neurotrophic factor release from the ARPE-19/BDNF EC-devices was considerably lower than the neurotrophic factor release from the ARPE-19/GDNF EC-devices.
Morphology
It has been hypothesized that the electrical responsiveness of the inner ear is closely related to the number of remaining SGNs in the Rosenthal´s canal. To investigate the amount of remaining SGNs after 4 weeks of EC device treatment, the SGN density in the Rosenthal´s canal was calculated using the Zeiss ZEN software Image. At the same time SGN areas were measured. The group that received ARPE-19/GDNF treatment had a significantly larger amount of remaining SGNs than the groups that received ARPE-19/BDNF, ARPE-19 or the empty devices.
WORK PACKAGE 4 – LONG-TERM SAFETY AND FUNCTIONALITY IN LARGE ANIMAL MODELS
Generating deafened animal model (Verena Scheper, Wiebke Konerding, Jana Schwieger)
All relevant SOPs and protocols for in vivo cat experiments were prepared and rapidly received an ethical approval. Within thr first project year, we received the permission to work on n=9 cats for the pilot trial. These animals have been deafened and implantated with an EC device as well as a cochlear-implant.
Pilot in vivo tests for proof of feasibility of the novel biodelivery devices (Verena Scheper, Wiebke Konerding, Jana Schwieger, Jens Tornøe, Jenny Ekberg, Chris Thanos, Briannan Bintz, Pavel Mistrik, Anandhan Dhanasingh, Jochen Tillein, Raphael Kiran)
Within the first project year we applied for the ethical approval for the pilot trial and received the permission for n=9 cats to be chronically deafened and implanted as pilot animals. Next to this, we performed temporal bone studies on EC improvement. Cat cadavers were co-implanted with EC devices designed for the cat cochlea by Partner 1 (NsGene, Denmark) and cat CIs for evaluation of optimal EC design. Cadaver experiments were performed with Dummy ECs of distinct diameters which were co-implanted together with CIs into cat scala tympani via the round window. Based on those experiments, we found the optimal design of the cat EC device to be 4 mm length and 0.4 mm diameter with a tether on one end for better handling.
The experimental animals were chronically deafened using daily subcutaneous injections of the aminoglycoside neomycin sulphate (60 mg/kg BW s.c.). So far, six animals have been included into the pilot trial. The first cat was deafened starting from postnatal day 14 after an evaluation of normal hearing status via AABR. Hearing status was assessed after 14 days (and 19 and 24 if necessary) of daily injections. As this cat did not show complete hearing loss after the maximal duration of chronic deafening, we decided, based on literature survey and discussions with other working groups, to switch to a neonatal deafening procedure starting on postnatal day 1 to 3. All of the following five animals lost their hearing ability within 24 days after birth/start of deafening procedure. None of the deafened animals showed clinically relevant signs of kidney failure, which can be a negative effect of aminoglycoside treatment. From the first three pilot cats blood samples were taken on the day of implantation and analyzed for kidney parameters. No relevant changes have been observed. The kidneys of the first three pilot animals were investigated by a pathologist for kidney anomalies and the histology did not show any pathological changes. Therefore we conclude that the neonatal deafening procedure is leading to a robust deafening which allows for further processing within this trial and that the deafening does not have relevant negative effects which may lead to a premature termination of experimental animals.
The five pilot animals underwent surgery for chronic co-implantation of the EC device and CI into the scala tympani. One subject died during surgery due to anesthesia failure. The other four received EC devices containing GDNF (n=3) or BDNF (n=1) producing ARPE-19 cells. In all implanted animals the handling of the EC devices allowed for an unproblematic co-insertion of the device into the scala tympani implanted with a cat CI. The design and size of the EC devices does not have to be further improved for in vivo cat implantation.
In all four implanted animals electrode impedances were measured directly after implantation, showing that the CI was intact. Following implantation, daily impedance measurements (via DibBox and Maestro Software from Med-El) were performed. During the first two weeks all electrodes exhibited high impedances indicating wire breakage and / or contacts being in non-conductive surroundings (i.e. being out of the cochlea). X-ray images were taken, showing that in all cases except the last implanted cat, the electrode had moved out of the cochlea. Based on this finding and due to the fact that the evaluation of all explanted implants showed wire breakage, the CI design was changed from experiment to experiment to eliminate all possible negative influences on wires and electrode placement. The cat CI development was led by Partner 3 (Med-El, Austria). Additionally, the surgical technique has continuously been improved based on the results and in accordance to suggestions from other working groups.
Two different approaches of CI implantation have been assessed: a subcutaneous and a transcutaneous approach. The subcutaneous electrode is similar to the human CI electrode in that all parts are implanted into the subject and coupling to an external stimulator is done via a subcutaneous coil. In the transcutaneous approach, which is the common approach in animal research, the electrode ends in a connector, which is reaching through the skin. Both approaches will be further explored to develop the best CI electrode for long-term chronic electrical stimulation in the cat.
Chronic preclinical tests in deafened cats for determination of long-term safety and functionality (Verena Scheper, Wiebke Konerding, Heike Janssen, Peter Hubka, Andrej Kral, Thomas Lenarz, Jenny Ekberg, Lars Wahlberg, Jens Tornøe)
In vivo testing of combined EC and CI implantation
The in vivo tests aimed to investigate the structural and functional effects of the selected factors released by encapsulated cells and devices produced within WP1 and WP2, using an animal model mimicking a severely deaf patient fitted with a cochlear implant. It is hypothesized that the electrical responsiveness of the auditory system will be significantly higher in animals treated with factor-releasing encapsulated cells.
The aims of the second half of the work package 4 schedule were:
1. To evaluate long-term biocompatibility and neuronal preservation using histological methods
2. Proof of long-term stability and functionality of EC-devices in vivo
All chronically implanted (CI and EC device) animals underwent the following procedures: deafening, training, implantation to the left ear, chronic measurements of CI functionality (impedances and hearing/auditory brainstem response threshold), acute CI implantation of left and right ear, final recording and histology of the cochlea.
The chronic implantation and electrical stimulation has been described in detail in Deliverable 4.3. In short, 21 cats were unilaterally co-implanted with an encapsulated cell (EC) device and a cochlear implant (CI) into the left scala tympani via the round window. The cats were assigned to three groups which received different treatment combinations of EC devices secreting glial cell-line derived neurotrophic factor (GDNF) and CI electrical stimulation (ES). Additionally, normal hearing animals served as controls.
The electrical stimulation of the +GDNF/+ES group (N = 7) started 14 days post-surgery. Cats were stimulated by passive acoustic stimulation (acoustic environment of the animal facility and additional radio broadcast) via the OPUS speech processor (MED-EL) and subcutaneous PULSAR implant (MED-EL) for 4 hours a day, 5 days a week. After confirming the functionality of the implant via electrically evoked auditory brainstem response (eABR) measurements, the level of stimulation was defined based on behavioural hearing thresholds. One animal was finalized already after 2 months, due to health problems unrelated to the cochlear implants. Thus, results are based on 20 animals (+GDNF/+ES: N =6, +GDNF: N = 8, -GDNF: N = 6) implanted for 6 months.
At the end of the 6-month implantation, we performed eABR measurements to assess the functionality of the remaining spiral ganglion neurons (SGN). For analysis (Matlab and R software) artifacts were blanked (1 ms) and signals filtered (300-5000 Hz).
To assess the overall response strength, we calculated root mean square (RMS) values (1-7 ms post stimulus interval). The input-output function (dependency of RMS value on stimulation level) was fitted with a sigmoidal function (formula 1):
y=a/(1+e^(-b(x-c)) )
where, a=upper asymptote (i.e. RMSMAX) [µV], b=slope [µV/dB], c=inflection point [dB]
The fitting procedure was successful in 14 left (implanted) and 18 right (untreated control) ears. Threshold levels (dB att.) were assessed at 10% RMSMAX and dynamic range (dB att.) was calculated from 10-90 % RMSMAX. The slope (µV/dB) was the maximal slope at the inflection point of the curve. Individual measurements were excluded when the fitted inflection point fell outside the stimulated dB-range.
To assess which parameter was best suited to represent differences in auditory nerve function, we first compared the untreated (right) ears to normal hearing controls. There was no significant difference between hearing controls (N=4) and neomycin deafened animals (untreated ear, N=18) for threshold, slope, or dynamic range. However, RMSMAX was significantly higher in controls than in neonatally deafened animals (meanhearing: 4.47 µV, SDhearing: 1.20 µV, meandeafened: 1.30 µV, SDdeafened: 0.57 µV; Mann-Whitney U test: p=0.003).
For the analysis of treatment effects, we calculated the relative percentile difference in RMS value between treated (left) and untreated (right) ear (formula 2): (left-right)/(right)∙100
Positive values indicated higher values in the treated compared to the control ear. The RMSMAX values were similar between ears, in all three groups (Wilcoxon signed rank test: control: p=0.125 +GDNF: p=1.000 +GDNF/+ES: p=0.625). Still, there was a statistical trend for a positive correlation between ES duration (+GDNF/+ES group) and RMSMAX (r=0.900 p=0.083) with longer chronic ES leading to higher RMS magnitudes at the left (treated) compared to the right ear.
The histological evaluation of the auditory nerve (neuronal preservation) as well as the EC device and fibrous tissue formation (biocompatibility) was described in detail in deliverable 4.4. In short, the animals were transcardially perfused under deep anesthesia and the cochleae were harvested and prepared for plastic embedding. To assess SGN survival, 5 midmodiolar slices at 20 µm distance (i.e. every 5th slice) were analyzed. In each, 4 sections of the Rosenthal’s canal were assessed: lower basal, upper basal, lower middle and upper middle. We calculated median SGN densities per mm2, based on the diameter of the Rosenthal’s canal and the number of SGN with clearly identified nucleus (LAS software, Leica microscope DM1000 & camera DFC 320, Leica GmbH). To assess the time course of SGN degeneration after neomycin deafening, SGN densities of the right (untreated) ear were compared to pilot animals (N = 6) and hearing controls (N = 4). The SGN density significantly declined with duration of deafness (Spearman correlation: r=-0.621 p=0.0003). The data was fitted with an exponential decay (GraphPad Prism, r2=0.932) yielding an Y0 (i.e. hearing controls) of 2119 cells/mm2, declining to a plateau of 237 cells/mm2 after about 8 months of neonatal neomycin deafening, with a half-life of 1.27 months.
We calculated relative percentile left-right differences between ears for SGN density (cf. formula 2), to assess the influence of treatment on auditory nerve survival. We separately assessed basal (median of lower and upper basal turn) and medial (median of lower and upper medial turn) parts of the cochlea. Positive values indicated higher SGN densities on the treated compared to the untreated ear. At the basal part of the cochlea was a significant difference between groups (Kruskal-Wallis test: p=0.0015) with +GDNF/+ES having significantly higher relative left-right difference in SGN density than the +GDNF group (Dunn’s Multiple Comparison test). Control animals and –GDNF animals had similar SGN densities at left and right ears, with relative left-right differences being not significantly different from zero (Wilcoxon signed rank test: control: p=0.0875 –GNDF: p=0.094). The +GDNF group showed a significant shift to negative values (p=0.023) indicating a lower SGN density on the left (treated) ear compared to the internal control ear. The +GDNF/+ES group, however, had significantly higher SGN densities at the left than right ear (p=0.031).
To assess the functionality of the EC device we took two approaches: 1) GDNF production assessment ex vivo and 2) histological assessment of the encapsulated cells in situ (for details see deliverable 4.4). Preimplantation concentration measured via enzyme-linked immunosorbent assay (ELISA) revealed mean production levels of 79.9 ng/24h/device (SD: 50.6 ng/24h/device, N=14). In the 6 EC devises remaining in situ for histological analysis (1 device from –GDNF group), we either found no ARPE-19 cells or cell debris (i.e. without cell nuclei; mean score: 0.7 range: 0-1). Of the EC devices producing GDNF, 9 were explanted at the end of the chronic implantation and were retested for neurotrophic factor release. Of these devices one was found to produce GDNF (animal 1421). This production (14.1 ng/device/day) was reduced to about 13% of the pre-implantation level (110.9 ng/device/day). Taken together, we showed directly (histology) or indirectly (ELISA) that ARPE-19 cells in 14 out of 15 EC devices died during a 6-month implantation time in the scala tympani of the cat. The cat with the EC device showing some remaining GDNF production (#1421) was no outlier in its group (+GDNF+ES), both concerning SGN survival and eABR measurements.
All devices left in situ were found to be encapsulated with a fibrous tissue sheath; either loose or dense (score 4 or 5). The mean score for fibrous tissue was 4.7 (SD: 0.4) with similar values throughout the cochlea (mean+SD: before: 4.7+0.5 midmodiolar: 4.8+0.4 after: 4.7+0.4). We concluded that this encapsulation was restricting diffusion of nutrients and was the most probable cause for the observed ARPE-cell death.
To assess potential effects of the GDNF treatment on fibrous tissue growth, we compared fibrous tissue in the scala tympani between left (treated) and right ear. At the right ear, only small amounts of fibrous tissue were found at the lower basal turn (median: 1, range: 0-1) and none was found in the upper basal turn. This was true for all individuals. At the left ear, we found on average higher amounts of fibrosis. For the –GDNF group a median of 1.5 (range: 0-3), for the +GDNF+ES group 1.5 (range: 1-3) and for the +GDNF group a median of 3 (range: 1-4). The left-right difference in fibrosis was not significant from zero for both the –GDNF and +GDNF+ES group (Wilcoxon signed rank test: –GDNF: p=0.149 +GDNF+ES: p=0.095) indicating similar levels of fibrosis at both ears. The distribution of the +GDNF group was significantly shifted to positive values, indicating a higher fibrous tissue growth at the left compared to the right scala tympani.
We propose that chronic ES was reducing the fibrous tissue amount in the scala tympani, allowing for an initial neuroprotective effect via GDNF delivered by the EC device. We conclude that additional anti-inflammatory treatments and/or changes of the EC device membrane would be feasible to reduce fibrosis and thus allow for a stable long-term deliverable of NTFs via EC devices in CI patients. Although the device was not stable over the 6 months implantation time in the deafened cat cochlea, the potent neuroprotective effect in the +GDNF+ES group clearly marks the potential of EC device treatment for human CI patients.
Potential Impact:
POTENTIAL IMPACTS
The NeuEar project aimed at developing genetically modified cells for encapsulation and long-term expression of selected neurotrophin(s) in a medical device implant prototype that can be combined with a cochlear implant. The NeuEar partners have also optimized safety and functional effects in in vitro and in vivo assays and evaluated the long-term safety and efficacy in animal models of a clinically relevant implant. The consortium has therefore successfully reached its final goal to have an implant prototype and functional data in suitable animal models available that support a continuation towards clinical testing after the NeuEar project has been finished.
Innovative approach
The implantable device represents a true potential for clinical application and future commercialization. This project aimed to be innovative but at the same time drew upon the previous successful technology development by the industrial partners (NsGene & MED-EL) and the research and translational experience of the academic partners (KI and MHH).
Several innovative developments have been brought forward by the project:
• First cochlear implantations of Partner 1’s EC device in guinea pig and cat deafness models as well as human temporal bone, expanding the scope of the technology platform.
• First combination of a cochlear implant with an encapsulated cell device releasing neurotrophic factors.
• Development of new cochlear implants for chronic implantation in cats.
• Development of the EPS concept allowing separate implantations of devices parallel to a cochlear implant (both human and animal).
Promotion of SMEs in Europe
The project, led by the SME NsGene, Denmark, integrates an industrial partner and two academic groups working together to reach the delivery of a therapeutic cell-based encapsulation device that can be implemented in conjunction with a cochlear implant. The consortium has demonstrated that it possessed all the expertise and know-how to develop, design, and validate this chronic implant and bring it close to the market.
Impact for the patients and the society
In Europe overall, more than 80 million persons (around 16%) suffer from hearing loss that is severe enough to adversely affect their daily life and only one in nine persons afflicted with hearing loss is still working. Despite common perception, hearing loss doesn’t only affect the elderly. The reality is that the majority (65%) of people affected by hearing loss are younger than age 65. In the States, 1.4 million children (18 or younger) have hearing problems; and it is estimated that 3 in 1000 infants are born with serious to profound hearing loss (survey from Better Hearing Institute).
Economic benefits
Untreated hearing loss costs Europe 213 billion euro per year, according to the international scientific report ”Evaluation of the Social and Economic Costs of Hearing Impairment”. This equals about 473 euro per year for each adult European. Of course, with an increase of the aging population since many decades, it is foreseen that the current prevalence of at least 3 in 10 people over age 60 having hearing loss will worsen. The great medical need has created a large worldwide market for cochlear implants, which was $725 million in 2008 and is expected to grow to $1.59 billion in 2014 (Neuroreports). Even though the current commercially-available implants restore hearing to great extent, their performance remain suboptimal under difficult hearing conditions (for instance in noisy environments). Furthermore, the established and possibly ongoing degeneration of primary auditory neurons (ANs) is a limiting factor in cochlear implant efficacy. The exogenous application of neurotrophic protein factors can prevent these degenerative processes and is expected to allow a degree a functional improvement. By combining a highly innovative device that will permit the long-term delivery of such neurotrophic factors with a cochlear implant, the project has a high socio-economic value for the European ageing society.
MAIN DISSEMINATION ACTIVITIES
NeuEar teams have placed a particular emphasis throughout the grant period to disseminate their work and scientific highlights to the scientific community, the European Commission, patient associations and public at large. You will find here a summary of the different actions taken:
Dissemination to the Scientific Community
Participation to international conferences, poster presentations and lectures
The full list of national and international conference attendance by the NeuEar teams is available on the ECAS portal.
List of publications
The full list of publications is available on the ECAS portal.
Creation of a NeuEar group on LinkedIn
To increase the visibilty of the NeuEar project within the scientific community, Isabelle Weiss (DWC, Partner 5) has created a NeuEar group on LinkedIn. The link to the NeuEar’s LinkedIn group is available on the NeuEar homepage.
Link: http://www.linkedin.com/groups/NeuEar-4843435?gid=4843435&trk=hb_side_g
22 members including 13 non-NeuEar members are following the NeuEar group on LinkedIn.
Publication of articles on the NeuEar project
• Lars Wahlberg and Jens Tornøe (Partner 1) and Isabelle Weiss (Partner 5) have been invited to write up an article on the NeaEar’s project and objectives by Pan European Networks (www.paneuropeannetworks.com). This jointly written article has been published in June 2013 in the issue 07 of PAN’s Science and Technology Journal (pages 210-211).
• Interviews given by Thomas Lenarz and Verena Scheper (Partner 4), and Jens Tornøe (Partner 1) to International Innovation were published in an article entitled ‘Loud and Clear’ on July 2014.
Development of a flyer to advertise the NeuEar’s research project
Partner 5 (DWC Ltd) in collaboration with Partner 1 (NsGene A/S) has rapidly developed a flyer for NeuEar. The flyer is presenting the research objectives of NeuEar, the teams involved and is acknowledging the EC funding. The flyer was developed in March 2013 to be used as a dissemination tool (see actions in the sections below). This flyer has been translated in French to be distributed to all kids who participated to the NeuEar outreach project in Spring 2015 (see section below).
Dissemination to Members of the European Parliament and the European Commision
The NeuEar flyer has been sent early 2013 to 45 Members of European Parliament together with the first NeuEar press release (published in February 2013).
Publication of press releases
A first press release has been published in February 2013. This press release was presenting the NeuEar scientific project and partners to the European Commission and the scientific community. By placing the link on the NeuEar website, it is also available to the public.
This press release was also published on Cordis – Weblink:
http://cordis.europa.eu/project/rcn/104052_en.html
Publication of the NeuEar’s project goals and objectives on the EC portal for Research and Innovation:
http://ec.europa.eu/research/health/medical-research/severe-chronic-diseases/projects/neuear_en.html
A second press release has been published in April 2015 following the annual meeting of the consortium in Innsbruck on March 2015. This press release was published on the NeuEar website and published on Cordis: http://cordis.europa.eu/news/rcn/123975_en.html
Dissemination to Patient Groups and Charities
In order to communicate with the relevant patients associations and charities that support European citizens impacted by hearing loss, Partner 5 (Isabelle Weiss from DWC) has listed 22 patient organisations throughout Europe using contacts already taken by the PIs of NeuEar and a web search. It has been indeed highly important for NeuEar to take and maintain contact with people suffering from hearing loss to inform them about the goals of the project with the development of this new device and to keep them informed on the research and technology progresses.
These 22 patient organisations listed have been contacted for the first time in March 2013 through a letter by Isabelle Weiss (Partner #5) to inform them about the scientific objectives of the consortium and to present them the first press release and the NeuEar flyer. All associations having their website listed on the NeuEar public website have been informed and did agree to it.
In return, some organisations have placed information on the NeuEar project on their own websites. Some examples are given below:
http://www.observatorioplastico.com/detalle_noticia.php?no_id=232511&seccion=envase&id_categoria=914
http://www.ingenieros.es/noticias/ver/proyecto-neuear-desarrollo-de-implantes-para-restaurar-la-funcion-auditiva/3408
These organisations are kept informed on the progress of the NeuEar project on a regular basis.
Dissemination to the public at large
NeuEar has developed early 2015 an ‘Art & Science’ outreach workshop dedicated to kids from 7 to 14 years old. Three workshops were presented in ‘Youth and Culture Centers’ (Maison des Jeunes et de la Culture in french) in Lyon and Bron (France) in April and May 2015 with a total of 45 children.
The goal of this 2-hour workshop was to intrigue and engage children with the wonders of the human ear and today’s technological developments like the cochlear implant. It was also to sensitize them to the different types of deafness and their origins, and to remind them about the importance of protecting their ears. For this workshop, we have developed several types of hands-on sessions together with a powerpoint document full of videos and pictures for a lively and fun presentation.
During the workshop, we have asked the children to locate the tympanum, the ossicles and the cochlea on a drawing presenting the 3 parts of the ear (external, middle and internal ear). Each section was given on a piece of paper to glue on the general drawing (see Picture 1). We have taken the opportunity of this hands-on session to describe the roles of the different sections of the ear and to explain the propagation of the sound through the ear and up to the brain. We have also explained what was the sound and its characteristics (frequency, decibels).
Kids were then asked to place different sources of noise from their everyday life on a decibel ladder; such as a washing machine, a vacuum cleaner, a classroom, a space rocket, a live concert, etc... This exercise triggered many discussions and questions in groups. Following this, we have created a simplified brain in air-dough to locate the brain region treating the sounds. The air-dough used for this hands-on session is a very light type of playdough drying to solid when left outside the plastic bag. This air-dough was colored so the kids could shape each cerebral lobe with a different color. Using this creative workshop, we have approached the notion of interpretation of sound and ear cell degeneration.
At the very end of the workshop, Dr Fabien Seldran from MEDEL, France presented a replica of a Medel cochlear implant to the children and explained the cochlear implant technology. 3-D model of the inner ear and cochlear implant insertion were also presented to the children to support our explanations.
Each child returned home with her/his own drawing of inner ear, her/his created model of brain, the NeuEar brochure translated in French and a flyer to help her/him creating new brain models at home. Both the NeuEar brochure and the workshop flyer have the NeuEar and EU logos. Hopefully these documents have triggered discussions at home with the parents!
EXPLOITATION OF RESULTS
Exploitation plan
NsGene has successfully brought the EC implant technology from research into clinical development for Alzheimer’s disease. In this process, the SME has gained insight and expertise in the necessary regulatory, manufacturing, and product development process. NsGene has therefore the ability to move the project beyond the 3 years of the initial phase. It is the intention of the consortium to apply for Phase Ib clinical testing based on successful completion of this 3-year project. If the EC implant is used in conjunction with a cochlear electrode, it is anticipated that the commercialization will be done in a partnership with MED-EL as this industrial partner has already established a commercially successful product in this therapeutic domain. The EC implant will also form a new product platform for regenerative therapies to the inner ear and as such could be developed into several pertinent products over the next few years.
List of Websites:
Dr Lars Wahlberg, COO, NsGene A/S
Tel: +45 2360 3198
E-mail: luw@nsgene.dk
