, Thomas Schanze2, Marcus Wilms2, Andreas Rentzos2, Ursula Brunner1, Florian Gekeler3 and Lutz Hesse4, 5| Graefe's Archive for Clinical and Experimental Ophthalmology Incorporating German Journal of Ophthalmology |
| © Springer-Verlag 2004 |
| 10.1007/s00417-004-1049-x |
Helmut G. Sachs1 , Thomas Schanze2, Marcus Wilms2, Andreas Rentzos2, Ursula Brunner1, Florian Gekeler3 and Lutz Hesse4, 5
| (1) | University Eye Clinic, University of Regensburg, Franz Josef Strauß Allee 11, 93042 Regensburg, Germany |
| (2) | Applied Physics-NeuroPhysics Group, Department of Physics, Philipps University Marburg, Marburg, Germany |
| (3) | Department of Neuroophthalmology, University Eye Hospital, University of Tuebingen, Tuebingen, Germany |
| (4) | Department of Ophthalmology, Philipps University Marburg, Marburg, Germany |
| (5) | Klinikum am Gesundbrunnen-Augenklinik, SLK-Kliniken Heilbronn GmbH, Heilbronn, Germany |
|
Helmut G. Sachs Email: sachs@eye-regensburg.de Phone: +49-941-9449205 Fax: +49-941-9449281 |
Received: 10 March 2004 Revised: 8 August 2004 Accepted: 13 September 2004 Published online: 1 December 2004
Millions of people worldwide suffer photoreceptor loss due to degenerative retinal diseases [20]. In Germany alone, 17,000 people become blind every year [22], and to date there is no effective treatment or cure. About 50% of all cases of blindness are caused by damage to the retina [22], and this population includes patients with macular degeneration and retinitis pigmentosa—conditions that cause progressive degeneration of the outer retina.
A number of approaches, including gene therapy and pharmacological measures, are currently being pursued in the hope of preventing blindness [2, 29]. However, once vision is lost, only two approaches have shown promise for restoring vision: retinal transplantation [29] and bioelectronic visual prostheses [1, 3–9, 11–14, 16–18, 20, 21, 23–26, 28, 30–32, 34–37].
In this latter category, two kinds of retinal implants are under development. Epiretinal implants are designed to stimulate retinal ganglion cells with an electrode array that is implanted on top of the inner limiting membrane. Subretinal implants are inserted between the pigment epithelium layer and the outer layer of the retina and attempt to stimulate intact remaining retinal cells by photodiode-generated currents. Previous experiments have shown that microphotodiode arrays (MPDAs) do not deliver sufficient energy to stimulate the retina [35].
The objective of the present study was to implant polyimide film electrodes subretinally in anaesthetised cats and to record the intracortical responses evoked by subretinal electrical stimulation in order to draw conclusions concerning the feasibility of this approach.
All
surgical and electrophysiological procedures were performed in Marburg
with anaesthetised cats. The procedures complied with the guidelines of
the European Communities Council Directive (86/609/EEC) and were
approved by an official German Animal Care and Use Committee. In
addition, the 
Principles of laboratory animal care
(NIH publication No. 85-23, revised 1985), the OPRR Public Health
Service Policy on the Humane Care and Use of Laboratory Animals
(revised 1986), the U.S. Animal Welfare Act, as amended, and the
Association for Research in Vision and Ophthalmology (ARVO) guidelines
were followed.
m with a spacing of 100 m in a given row. Figure 1 illustrates the electrode design.
Three adult cats (body weight 3–4 kg) received atropine sulphate (0.03–0.05 mg/kg) to reduce salivation. Anaesthesia was induced by intramuscular injection of a mixture of ketamine hydrochloride (Ketanest, 10–20 mg/kg) and xylazine hydrochloride (Rompun, 1–2 mg/kg). During eye surgery the cats were fixed on their back in a modified Horsely–Clarke support and anaesthesia was maintained with ketamine hydrochloride (Ketanest, 1–10 mg/kg). Local anaesthetics (lidocaine, bupivacaine) were used in addition. The level of anaesthesia was controlled by monitoring body respiration, ECG, and reflexes. Body temperature was maintained at about 38°C.
clock
position between the access for the vitrectomy probe and necessary
intraocular instrumentation and the access for the stimulation film. No
fiberoptic intraocular light source was necessary to illuminate the
intraocular surgical field. The coaxial illumination from the operating
microscope falling through the noncontact wide-field lens system (BIOM
II, Ocular Instruments, California, USA) was bright enough to perform
the entire surgical procedure. This is due to the highly reflecting
tapetum lucidum in the cat. A very few disturbing light reflexes
appeared in the observation lens systems but these did not hamper
intraocular surgery, thus permitting bimanual handling of
instrumentation and film. The vitrectomy was carried out using standard
surgical parameters. Only the nasal and anterior parts of the vitreous
cavity are difficult to access through this lateral port without
touching the lens. Next the retina was prepared for film implantation:
a retinotomy was performed in an inferior temporal mid-peripheral area
by injecting balanced salt solution (BSS) via a Teflon cannula into the
subretinal space and conventionally opening the retina with vitreous
scissors in the bleb area thus created. A viscoelastic substance
(Healon 0.1–0.3 ml) was introduced via the retinotomy into the
subretinal space to ease the subretinal introduction of the stimulation
film from the vitreous cavity. To stiffen the highly flexible
stimulation film, two more rigid guide films similar in shape to the
stimulation film were used during introduction into the vitreous
cavity. The stimulation film was sandwiched between the guide films
during scleral passage. The guide films were withdrawn when the
stimulation film in the premacular region was grasped with vitreous
forceps. With the aid of these forceps the stimulation film was
advanced through the retinotomy into the subretinal space as far as the
desired macular area (Fig. 2). The inferior
temporal mid-peripheral area for the retinotomy was chosen
corresponding to the sclerostomy for the foil implantation. This
allowed placement of most of the electrodes, which are distributed in
rows of logarithmic ascending distances on the stimulation film, in the
central posterior area by advancing the foil through the retinotomy.
The viscoelastic solution and subretinal fluid in the bleb area were
aspirated and PFCL was injected carefully. 7.0 Vicryl was used to
suture the extraocular film onto the sclera and to seal the
sclerostomies. The cats were then repositioned for subretinal
electrical stimulation and simultaneous cortical recording. Therefore,
the globes were filled with PFCL to a maximum extent. Repositioning has
to be performed with extreme care to avoid any dislocation of the
subretinal film due to movement of the PFCL. The cats were repositioned
in a way that the PFCL was covering the implanted stimulation film
completely at any time during this manoeuvre. Film position was then
checked ophthalmoscopically prior to the stimulation experiments and
photodocumentation was performed with a handheld fundus camera (KOWA,
Kowa Ltd, Tokyo, Japan). An standard exchange of PFCL against silicone
oil was carried out after the recording session. The reopened
sclerostomies were closed again and the extraocular foil portion was
hidden under the conjunctiva. The retinotomies and the accidental
retinal perforation were not treated by retinopexy. The operated eyes
were controlled according clinical demands in a follow-up period of
3 months ophthalmoscopically.
For cortical recording the cats were positioned in a standard Horsely–Clarke support. Anaesthesia was maintained by ventilation with N2O/O2 (70%/30%) and halothane (0.3–0.8%) (n=1) or isoflurane (0.5–1.5%) (n=2). Continuous monitoring of rectal temperature (38°C), end-expiratory CO2 (3.8–4.2%), ECG, EEG, and reflexes was used to control the level of anaesthesia. In addition to general anaesthesia, local anaesthetics (lidocaine) were applied during the skull surgery required to insert up to seven fibre electrodes into the primary visual cortex.
Electrical
stimulation was performed in three cats with computer-generated
sequences of voltage waveforms and fast voltage–current converters with
zero offset current [27]. The applied stimulation currents consisted of short charge-balanced rectangular impulses (400 s) with amplitudes ranging from 10 to 100 A.
The currents were applied by the subretinally placed film electrodes
against a distant large reference electrode located at the rectal body
temperature sensor.
Cortical recordings in
the primary visual cortex were performed with up to seven
quartz-isolated platinum–tungsten electrodes (80 m diameter) placed in locations corresponding to retinal stimulation sites using an Eckhorn electrode drive [10].
Three types of signals were extracted: (1) single unit activity; (2)
multiple unit activity; and (3) local field potentials. Alternatively,
broadband cortical signals were band-passed (1–4000 Hz) and stored on
hard disk/CD for subsequent data analysis. For details see ref. [27].
Averages of the cortical responses evoked by focal electrical retinal stimulation were computed in order to assess electrical retinal stimulation. Cortical activation distributions were then computed according to the method described by Schanze et al. [27]. To estimate the temporal resolution of cortical responses to retinal stimulation by a waveform approach, all values were renormalised to a retinal eccentricity of 2° visual angle. In this case the duration of the first cortical excitatory deflection was estimated.
All three implantations were performed successfully. In all three feline left eyes operated, the film electrodes were completely implanted subretinally in the area centralis.
By comparison with human eyes, the feline retina is considerably more difficult to detach surgically. The partial detachment required for subretinal placement of the stimulation film was achieved in all three eyes. Manipulation of the subretinal film for submacular placement caused unintentional perforation of the retina out of the subretinal space in one animal. Satisfactory submacular placement was achieved in this case by retracting and readvancing the film. Once the film is in a subretinal position and has been stabilised with PFCL, the extraocular part of the stimulation film can be carefully fixed to the sclera without causing dislocation. This permits further manipulation of extraocular film components during the stimulation procedure, such as connecting the film to the stimulation unit with a microplug. During all three experiments neither displacement of the film nor any other adverse events, especially inflammatory reactions, were detected over a 12-h period. At the end of the experiments the stimulation unit was disconnected and the films were mobilised and withdrawn. PFCL was removed and the sclerostomy, conjunctiva and canthotomy were sealed. All cats regained consciousness without problems within 2 h after skull closure. No adverse events were detected for the operated eyes during the 3-month follow-up period.
A were detected for biphasic charge-balanced stimulation currents (Fig. 3).
The excitatory cortical group responses peaked at about 20 ms following
electrical stimulation. As expected, the distribution of the cortical
responses was retinotopically related to retinal stimulation sites.
Estimation of the average full width at half height of the cortical
activation distributions yielded about 1 mm cortex, corresponding to 1°
visual angle. The temporal resolutions ranged from 20 ms to 50 ms.
The ocular surgery for implanting subretinal film-bound stimulation arrays in feline eyes was developed by applying surgical techniques and instruments normally used in humans for routine pars plana vitrectomy and buckle surgery.
Low-threshold stimulation is important for the safety of retinal tissue. Stimulation thresholds around 50 A
were measured in our experiments approximately 2 to 5 h after retinal
electrode implantation. The surgical trauma might be an explanation for
the relative high thresholds of around 50 A.
A successful retinal stimulation may have a lower threshold in an
intact retina or might be assumed if the retina were allowed to recover
for a longer period. The observed differences in the animals may be
caused by different retinal conditions shortly after surgery and seems
to be worth looking at in further experiments. The calculated charge
densities were approx. 2 mC/cm2 for the platinum-coated titanium nitrite electrode with its fractal surface geometry and are therefore in a reasonable range.
The measured spatial and temporal resolutions of cortical responses evoked by subretinal electrical stimulation were in the same range as those reported for epiretinal stimulation with fibre or polyimide film electrodes [16, 27].
These experiments have demonstrated the feasibility of film-bound subretinal electrical stimulation. The model suggests a variety of possible further experiments to determine electrical and technical parameters in subretinal visual prosthesis development. Because of the accumulated wealth of knowledge concerning the feline visual system and the availability of the model, it appears reasonable to use this experimental design to collect extensive and essential data in preparation for electrical subretinal stimulation experiments in humans.
On anatomical grounds, the cat may not be the ideal model for the development of surgical techniques in connection with subretinal implants. However, the feline model does provide extensive additional information if interest is focused not only on the surgical aspects but also on the desired function of a stable implant. Therefore the feline model should not be rejected in the context of visual prosthesis development.
Our results further indicate that short focal subretinal stimulation evokes localised spatio-temporal cortical responses. While coarse restoration of vision should be feasible with the subretinal implant, further experiments are necessary to determine the potential and limitations of this visual prosthesis approach.
Braille
reading by a blind volunteer by visual cortex stimulation. Nature 259:111–112