CURRENT STATUS OF REHABILITATION FOR PATIENTS WITH HOMONYMOUS FIELD DEFECTS Susanne Trauzettel-Klosinski MD Center for Ophthalmology, University of Tuebingen Germany

LEARNING OBJECTIVES INTRODUCTION 1) The att endee will be able to describe the disabiliti es In pati ents with brain damage, oft en the hemianopic and handicaps in everyday life caused by hemianopia. fi eld defect is not diagnosed, because other neurological symptoms, such as hemiplegia, are predominant. The 2) The att endee will be able to explain the features of the mean ti me between the brain damage and the diagnosis reading disorder and its rehabilitati on opti ons. of the hemianopia was found to be 3 +/- 2 months (Zhang 3) The att endee will be able to understand what et al 2006). Additi onally, the pati ents oft en do not realize rehabilitati on approach to improve orientati on is the fi eld defect, even though they experience acti vity appropriate. limitati ons in their everyday life.

The classifi cati on of the World Health Organizati on (WHO 2004) for Functi oning, Disability and Health (ICF) includes CME QUESTIONS three main aspects, which have to be considered in visual 1) Which spontaneous adapti ve strategies are favorable impairment: 1) impairment – related to the organ, 2) to improve orientati on? disability or acti vity limitati on – related to the person and 3) handicap or parti cipati on restricti on – related to the a) turning the head to the blind side society. b) saccades towards the blind side Homonymous fi eld defects cause two main disabiliti es/ c) shift ing att enti on to the blind side acti vity limitati ons: 2) The hemianopic reading disorder is caused by 1. reading disorder, if the central visual fi eld is involved a) low visual acuity 2. orientati on disorder (bumping into objects, problems b) fi xati on instability with way fi nding) c) limited size of the reading visual fi eld These disabiliti es cause major handicaps, which are listed in 3) Which rehabilitati on approach to improve the Table 1. hemianopic orientati on disorder is evidence-based? a) compensatory saccadic training Table 1: Handicaps resulti ng from hemianopia

b) visual sti mulati on of the blind hemifi eld  Reduced parti cipati on in society c) prisms mounted in the spectacles  Missing spati al informati on  No driving KEYWORDS  Decreased acti viti es of daily living 1. Homonymous Hemianopia  Impaired independence 2. Reading  Reduced social contact 3. Orientati on  Severe reducti on of quality of life! 4. Rehabilitati on

5. Training Because of these severe disabiliti es and handicaps, rehabilitati on is important and should be provided for these pati ents.

2011 Annual Meeting Syllabus | 91 General Aspects Of Rehabilitati on Fig. 1 A: Fixati onal eye movements during a strict fi xati on Three main aspects need to be considered: task. The fi xati onal eye movements are asymmetric towards the hemianopic side. B: By eccentric fi xati on and/ 1. Knowledge and potenti al uti lizati on of spontaneous or saccades towards the hemianopic side the verti cal adapti ve strategies visual fi eld border is shift ed towards the blind side, as well as the blind spot. The blind spot serves as a reference 2. Opti cal devices scotoma. This mechanism leads to a new perceptual along 3. Training the verti cal midline and is oft en misinterpreted as a real It is crucial for training studies to exclude spontaneous improvement of the visual fi eld. C top: With straight gaze, recovery. Additi onally, placebo eff ects have to be there is no informati on from the hemianopic side. C: ruled out by using a control group. Furthermore, it is With explorati ve saccades towards the hemianopic side necessary to defi ne appropriately, which improvement the visual fi eld defect is shift ed, resulti ng in a bett er use is clinically relevant and to assess the potenti al success of the fi eld of gaze. Obstacles, such as the suitcase and of a training by suitable methods. the person, can be detected in ti me. This strategy can be improved by saccadic training (Roth et al 2009) (modifi ed In principle, there are two approaches for training aft er Trauzett el-Klosinski 2010). methods: resti tuti on or compensati on (see below for details).

THE HEMIANOPIC ORIENTATION DISORDER OPTICAL DEVICES FOR BETTER ORIENTATION Pati ents with homonymous hemianopia are oft en not Monocular prisms or mirrors can enlarge the binocular aware of their fi eld defect and are confused by unpleasant fi eld. A benefi t was described in 20% of the pati ents “events” such as bumping into objects or persons, as well (Hedges et al., 1988). However, the competi ti on between as problems with way fi nding. The pati ents oft en develop the seeing halves of the reti na leads to confusion and spontaneous adapti ve strategies: eye movements towards impairment of spati al orientati on. A newer method uses the blind side (Huber et al 1995, Pambakian et al 2000, monocular peripheral prisms, where the prisms are Trauzett el-Klosinski and Reinhard 1998, Reinhard et al 2005) located only in the peripheral part of the glasses in order that are small during fi xati on tasks (see Figure 1 A) and to avoid the central diplopia (Peli, 2006). This approach larger during explorati on tasks, where they allow a bett er has been described to be quite successful: 47% (20 of 32) use of the fi eld of gaze (Figure 1 C). Furthermore, pati ents of the pati ents where wearing the prisms aft er 12 months can develop an att enti onal shift towards the blind side. It is and reported benefi ts for obstacle avoidance (Bowers et known that att enti on improves sti mulus discriminati on (Pilz al., 2008). However, this study was not randomized and et al 2006). Some pati ents get accustomed to a head turn, controlled, and the success was mainly determined by the which is unfavorable. Interesti ngly, the eye movements subjecti ve report of the pati ents. during fi xa ti on are asymmetric towards the blind side, thus causing a shift of the visual fi eld border towards the blind Binocular prisms have been shown to be benefi cial in or side. Huber et al (1995) found in a SLO-perimetry study in pati ents with hemineglect (Rose tti et al 1998). all 15 pati ents with hemianopia fi xati onal shift s: 12 pati ents with 1-5°, 3 pati ents with 5-15°. This shift of the visual fi eld TRAINING border is oft en misinterpreted as an improvement of the There are two approaches to improve the hemianopic visual fi eld. The exact determinati on of the positi on of the orientati on disorder: blind spot helps to identi fy whether fi xati on is central or 1. Resti tuti on by visual sti mulati on of the blind eccentric or whether eye movements shift the fi eld defect hemifi eld. together with the blind spot (Figure 1B). 2. Compensati on by increasing the fi eld of gaze with explorati ve saccades.

Previous studies performed visual sti mulati on using targets along the visual fi eld border. Zihl and von Cramon (1979) used targets at threshold and described an improvement of up to 40°. These results were not confi rmed by a later study by Balliet et al. (1985).

Some years later, Kasten et al. (1998) used supra-threshold targets along the visual fi eld border and recorded an improvement of 5°. The eff ect of this “Vision Resti tuti on Training VRT” of 5° improvement along the verti cal fi eld border was not confi rmed by Reinhard et al. (2005) using SLO perimetry, where fi xati on was simultaneously

92 | North American Neuro-Ophthalmology Society controlled during sti mulus presentati on. The test point grid peripheral sti mulus might be safer regarding provocati on had a spati al resoluti on of 0.5° horizontally and 1° verti cally. of saccades, but it is not clear whether eye movement Also in conventi onal perimetry using a threshold oriented arti facts can be excluded. slightly supraliminal stati c grid procedure (Tuebingen Automated Perimetry) no relevant change in the fi elds The alleged eff ects of the above menti oned resti tuti on was found (Schreiber et al 2006; same pati ent group as in studies should be disti nguished from the “blindsight” Reinhard`s study, Reinhard et al 2005). The problem with phenomenon, which is an unconscious percepti on of visual conventi onal perimetry is insuffi cient fi xati on control and sti muli via the superior colliculus to extrastriate regions the provocati on of eye movement towards the sti mulus. without acti vati on of V1 (Pöppel et al., 1973; Vanni et al., Furthermore, light scatt er of a bright sti mulus near the fi eld 2001; Weiskrantz, 2004). It was shown recently that the border may interfere. Another problem in the studies by thalamic lateral geniculate nucleus has a causal role in V1- Kasten and Sabel`s group is the method used to assess the independent processing of visual informati on (Schmid et al visual fi eld. 2010). Whether blindsight training can improve this kind of residual vision to a level that is relevant for everyday life, is Kasten and Sabel (1998) reported especially on an open questi on (see also Schofi eld and Leff 2009). improvement along the verti cal fi eld border, where they reported that absolute fi eld defects changed to relati ve. COMPENSATING SACCADIC TRAINING However, we need to consider that the so-called high Several studies reported improvement of effi ciency of resoluti on perimetry HRP in their studies did not measure explorati on aft er compensati ng saccadic training (Kerkhoff relati ve fi eld defects, but rather how oft en a sti mulus is et al., 1992; Zihl, 1995; Nelles et al., 2001; Pambakian et al., seen or not seen. It is typical for frequent shift of gaze 2004; Bolognini et al., 2005 - the latt er using crossmodal that at a certain locati on a sti mulus is someti mes seen informati on by including additi onal acousti c sti mulati on). and someti mes not seen. However, this is not the same However, none of these studies was randomized and as a relati ve defect. What they interpreted as relati ve controlled, and therefore the proof for the eff ect of a visual fi eld defect is in reality an eye movement arti fact. compensati ng saccadic training could not be shown. Therefore, it should be emphasized that the assessment of Therefore, we performed a randomized and controlled an improvement needs to be performed by an appropriate study, in which we compared explorati ve saccade training method. Additi onally, regarding the anatomy and with visual sti mulati on training in pati ents with hemianopia pathology, there is no reason why the visual fi eld should (Roth et al., 2009). improve just along the ver ti cal fi eld border (Horton, 2005). 30 pati ents with post-chiasmati c lesions were included. Therefore, the main limitati ons of the assessment of the The main inclusion criteria were: isolated homonymous VRT´s eff ect are as follows: 1. There is insuffi cient fi xati on hemianopia or quadrant defect, normal or near normal control, which cannot detect eye movements towards the visual acuity, and a durati on of the disease of at least 6 sti mulus. 2. “High resoluti on perimetry” does not measure months (to exclude spontaneous recovery, see Zhang relati ve fi eld defects, but the frequency of seen sti muli. et al., 2006). The main exclusion criteria were other 3. We need to consider the possibility of light scatt er by ophthalmological, neurological or cogniti ve disorders. The a sti mulus that is presented directly along the visual fi eld pati ents were randomly assigned: 15 to the explorati ve border, especially with unstable fi xati on. Additi onally, an saccadic training (EST) and 15 to a fl icker-sti mulati on improvement of 5° would not have a relevant eff ect on training (FT) – a potenti al resti tuti on training. the orientati on disorder in everyday life (Trobe et al 2005). For reading, however, this could have a crucial eff ect (see The EST consisted of a saccadic search task to improve below), but the pati ents in the study of Reinhard et al visual search in the blind hemifi eld and the use of the total (2005) did not improve their reading ability aft er VRT. fi eld of gaze. The training was performed on a computer. Positi ons and latency of all digits found were stored. The FT To argue against the possibility of eye movement arti facts, was designed in such a way that the pati ents had to fi xate later studies with very small sample size were performed a central panel and fl ickering lett ers were displayed in 22° with eye movement recordings during HRP (Kasten and eccentricity randomly on both sides. This kind of training Sabel 2006) and with camera-based microperimetry was thought to eliminate eye movements, whereas the (Marshall et al 2010). However, they cannot prove their saccadic training was designed to provoke eye movements. conclusions based on the data of these studies, because several technical and methodological details are not The pati ents trained at home, 30 minutes twice a day, clarifi ed. Furthermore, they do not show clinical relevance. fi ve days a week for six weeks (total 30 hours). Data collecti on was before, directly aft er and six weeks aft er Another study used supra-threshold targets at 10° training. The main results were: the digit search task at the eccentricity (Raninen et al., 2007). The authors reported computer showed a marked decrease of reacti on ti me in a normalizati on of contrast sensiti vity in the blind fi eld of EST compared to FT on the blind side. In a natural search two pati ents, but no visual fi eld improvement. The more task, a table test, where pati ents had to fi nd 20 objects

2011 Annual Meeting Syllabus | 93 (equally distributed in the total fi eld), the pati ents showed small paracentral homonymous fi eld defects can easily be a selecti ve decrease of reacti on ti me on the blind side overlooked in standard perimetry, especially when the test in EST. The eff ect to digit search and natural search task point grid is not dense enough to detect a scotoma of 2° sti ll persisted six weeks aft er the training at follow-up. diameter. In these cases, the examiner has to specially look During natural scene explorati on, the number of fi xati ons for such small defects by using a dense grid or a manual increased on the blind side and decreased on the seeing perimetry. side in EST, whereas in FT, there was no diff erence between the sides. The eff ect of increasing number of fi xati ons on The side of the fi eld defect in relati on to the reading the blind side even increased aft er the end of training, directi on plays an important role: In left to right reading, which shows that the pati ents applied their new strategy pati ents with right hemianopia have diffi culti es to get in everyday life. Quality of life improved in EST in the social through the line, make many small saccades and are domain, i.e. the pati ents detected other persons in their much more disabled compared with pati ents with left blind hemifi eld and could communicate with them. Reading hemianopia, who have diffi culti es to fi nd the beginning of speed did not change in either group, which was expected, the new line (see also TK 2010 JNO). because it was not a reading training, and neither did their visual fi elds. Spontaneous adapti ve strategies during reading: Some pati ents without macular sparing can develop a very To summarize the results of the randomized and controlled favorable adapti ve strategy: eccentric fi xati on (Trauzett el- trial (Roth et al 2009): saccadic training selecti vely Klosinski 1997). Even though they have normal visual improved saccadic behavior, natural search and natural acuity, they are able to shift their fi xati on by 1 - 1.5° to the scene explorati on. The pati ents learned to apply their side and therefore shift the visual fi eld border towards the new saccadic strategy to everyday life. On the other hand, blind side. By doing so, they create a small perceptual area fl icker training did not change the visual fi elds, explorati on along the verti cal fi eld border, which is very helpful for was unchanged, and fi xati on became more unstable enlarging the reading visual fi eld (Figure 2 B). On the other (unsystemati c eye movements). In conclusion, there is sti ll hand, one needs to consider that the eccentric fi xati on also no evidence for visual fi eld resti tuti on by sti mulati on of the causes a shift of the fi eld border in conventi onal perimetry, blind hemifi eld. together with the blind spot, and this behavior is oft en misinterpreted as an improvement of the visual fi eld, Compensati ng saccadic training is evidence-based and whereas in reality it is an adapti ve strategy (Figure 1 B). recommendable. Our study showed a clear eff ect, which could be applied to everyday tasks and which persisted Another adapti ve strategy is predicti ve saccades in left aft er the end of training. hemianopia, where the pati ents in the early stage make many hypometric saccades to fi nd the beginning of the new Interes ti ngly, pati ents who had a longstanding hemianopia line. Aft er some ti me they can learn to perform one single (for many years) also benefi ted from the training. hypermetric (predicti ve) saccade, and then start to read the new line (Meienberg 1988). Alternati ve methods: If the use of a computer is not possible, search tasks on a sheet of paper or a game with Reading performance in hemianopia therefore depends on search tasks (e.g. Domino) can be used (Kerkhoff et al 1992, several factors. Zihl et al 1995). Even though there are no randomized and 1. The side of the fi eld defect (right-sided hemianopic controlled studies of this kind of training, it is in principle pati ents are much more disabled). the same approach and could be applied in pati ents, who are too disabled or have additi onal cogniti ve or motor 2. The distance of the fi eld defect to the midline (i.e. problems. the size of the reading visual fi eld). 3. The presence and absence of adapti ve strategies THE HEMIANOPIC READING DISORDER such as eccentric fi xati on and predicti ve saccades. Reading requires not only suffi cient resoluti on, but also a suffi cient size of the reading visual fi eld or visual span during fi xati on (Aulhorn 1953, legge et al 1997, McConkie and Rayner 1975). A homonymous fi eld defect without macular sparing covers half of the reading visual fi eld and leads to severe reading disability (Figure 2 A). If there is a macular sparing, reading ability can be preserved (Figure 2C) even though the fi eld defect causes severe orientati on disability. On the other hand, a small paracentral homonymous fi eld defect results in severe reading disability, because it covers half of the reading visual fi eld (Figure 2 D), whereas orientati on is not involved. These

94 | North American Neuro-Ophthalmology Society CONCLUSIONS To improve explorati on, orientati on and mobility, saccadic training methods are recommendable. Training with a saccadic search task is now evidence-based and can be recommended for pati ents with hemianopia. The training was tested in a randomized and controlled trial (Roth et al 2009) and is available as soft ware (for further informati on see www.eye.uni-tuebingen.de/low-vision-clinic). Similar search tasks are used in other training methods (Zihl et al 1984, Kerkhoff et al 1992). The training, which was developed in our laboratory, has the advantage that it is designed in a very simple and clear way, so that even pati ents who have no experience with a computer, can use Fig.2 A: In macular splitti ng, half of the reading visual fi eld it easily and independently at home – aft er an instructi on is covered and reading is severely impaired. C: In macular by their ophthalmologist, neurologist or low vision sparing, the reading visual fi eld is preserved and reading specialist. Other training methods with explorati ve search can be normal, despite a large fi eld defect, which involves tasks use, in principle, the same approach. orientati on. D: In a homonymous paracentral scotoma half of the reading visual fi eld is covered and there is severe For reading, training with scrolled text can be reading disability, even though it is only a very small fi eld recommended (Spitzyna et al., 2007). Further methods for defect (approximately 2°). B: In eccentric fi xati on the improving reading need to be developed in the future. fi eld defect is shift ed towards the hemianopic side and a Altogether, evidence-based rehabilitati on procedures perceptual area is created to enlarge the reading visual fi eld. are available and should be provided for the pati ents to improve their quality of life.

HOW CAN READING BE IMPROVED IN HEMIANOPIA? CME ANSWERS - It is always important to support orientati on on 1 B,C the page by visual or tacti le aids (index fi nger, ruler or slightly magnifying ruler with a guide line). 2 C - Predicti ve saccades, especially in left hemianopia. 3 A - Turning the text to a verti cal or diagonal orientati on is oft en recommended, however there are no systemati c studies in pati ents and there are REFERENCES only anecdotal reports. 1. Aulhorn, E.: Über Fixati onsbreite und Fixati onsfrequenz beim Lesen - Computer training with scrolled text has been gerichteter Konturen, Pfl ügers Arch Physiol 257:318-328, 1953. described to be eff ecti ve (Zihl et al., 1984; Kerkhoff 2. Balliet, R., Blood, K.M., Bach-Y-Rita, P.: Visual fi eld rehabilitati on in the et al., 1992; Spitzyna et al., 2007). corti cally blind? J. Neurol. Neurosurg. Psychiatry 48:113-1124, 1985. - Systemati c oculomotor training without text (but 3. Bischoff , P., Lang, J., Huber, A.: Macular sparing as perimetric arti fact, with the size of reading saccades) has also been Am J Ophthalmol. 119(1):72-80, 1995. reported to be eff ecti ve (Schütt et al., 2008, 2009). 4. Bolognini,N., Rasi, F., Coccia, M. et al.: Visual search improvement in hemianopia pati ents aft er audio-visual sti mulati on, Brain 128:2830- Despite the reported benefi cial eff ects, more research 2842, 2005. to develop the opti mal training strategies for improving 5. Buonomano, D.V., Merzenich, M.M.: Corti cal plasti city: From reading performance will be necessary. synapses to maps, Av Rev Neurosci 2: 149-186, 1998. 6. Bowers, A.R., Keeney, K., Peli, E.: Community-Based Trial of a HOW SHOULD READING PERFORMANCE BE ASSESSED? Peripheral Prism Visual Field Expansion Device for Hemianopia, Arch Reading performance should be measured by standardized Ophthalmol 126(5):657-664, 2008. reading texts, especially paragraphs of texts, which are 7. Hedges, T.R. Jr., Stunkard, J., Twer, A.: Fresnel Prismen: ihr more favorable than single sentences. The Internati onal Stellenwert in der Rehabilitati on homonymer Hemianopsien, Klinische Monatsblätt er für Augenheilkunde 192:568, 1988. Reading Speed Texts (IReST), developed in 17 languages, provide a set of ten equivalent texts in each language 8. Hier D.B., Mondlock, J., Caplan, L.R.: Recovery of behavioral abnormaliti es aft er right hemisphere stroke, Neurology 33 for repeated measurements and also for the use in March:345-350, 1983. internati onal mul ti -language studies (Hahn et al., 2006, 9. Horton, J.C., Hoyt, W.F.: The representati on of the visual fi eld in Trauzett el-Klosinski et al, in preparati on), (for more human striate cortex: a revision of the classic Holmes map, Arch. informati on see www.amd-read.net). Ophthalmol. 109:816-824, 1991.

2011 Annual Meeting Syllabus | 95 10. Horton, J.C.: Disappointi ng results from Nova Vision`s visual 33. Roth, T., Sokolov, A.N., Messias, A. et al.: Comparing explorati ve restorati on therapy, Br J Ophthalmol 89:1-2, 2005. saccade and fl icker training in hemianopia: a randomized controlled study, Neurology 72(4):324-331, 2009. 11. Kasten, E., Wust, S., Behrens-Baumann, W.: Computer-based training for the treatment of parti al blindness, Nat Med. 4(9):1083-1087, 34. Rovamo, J. and Virsu, V.: An esti mati on and applicati on of the human 1998. corti cal magnifi cati on factor, Exp Brain Res 37(3):495-510, 1979. 12. Kasten, E., Bunzenthal, U., Sabel, B.: Visual fi eld recovery aft er vision 35. Schmid, M.C., Mrowka, S.W., Turchi, J. et al.: Blindsight depends on restorati on therapy (VRT) is independent of eye movements: an eye the lateral geniculate nucleus, Nature 466:373-377, 2010. tracker study, Behav Brain Res 175:18-26, 2006. 36. Schofi eld, T.M., Leff , A.P.: Rehabilitati on of hemianopia, Curr Opin 13. Kerkhoff , G., Münßinger, U., Haaf, E. et al.: Rehabilitati on of Neurol 22:36-40, 2009. homonymous scotoma in pati ents with postgeniculate damage of 37. Schreiber, A., Vonthein, R., Reinhard, J. et al.: Eff ect of visual the visual system: saccadic compensati on training, resti tuti on training.on absolute homonymous scotomas, Neurology Restorati ve Neurolology and Neuroscience 4:245-254, 1992. 67: 143-145, 2006. 14. Kerkhoff , G., Münßinger, U., Meier, E-K.: Neurovisual rehabilitati on in 38. Schuett , S., Heywood, C.A., Kentridge, R.W. et al.: Rehabilitati on of cerebral blindness, Arch Neurol 51 (5):474-481, 1994. hemianopic dyslexia: are words necessary for re-learning oculomotor 15. Kerkhoff , G.: Neurovisuelle Rehabilitati on nach Hirnschädigung, control? Brain 131:3156-3168, 2008. Ergotherapie 1(2):53-65 2000. 39. Schuett , S.: The rehabilitati on of hemianopic dyslexia, Nature 16. Kölmel, H.W.: Die homonymen Hemianopsien, Springer, Berlin, 1988. Reviews Neurology 5:427-437, 2009. 17. Legge, G.E., Ahn, S.J., Klitz, T.S. et al.: Psychophysics of reading - XVI. 40. Schweigart, G., Eysel, U.T.: Acti vity-dependent recepti ve fi eld changes The visual span in normal and low vision, Vision Research 37:1999- in the surround of adult cat visual cortex, Eur J Neurosci 15:1585- 2010, 1997. 1596, 2002. 18. Marshall, R.S., Chmayssani, M., O‘Brien, K.A. et al: Visual fi eld 41. Spitzyna, G., Wise, R., McDonald, S. et al.: Optokineti c therapy expansion aft er visual restorati on therapy, Clin Rehabil DOI: improves text reading in pati ents with hemianopic alexia, Neurology 10.1177/0269215510362323, 2010. 68:1922, 2007. 19. McConkie, G.W., Rayner, K.: The span of the eff ecti ve sti mulus during 42. Trauzett el – Klosinski, S.: Eccentric fi xati on in hemianopic fi eld a fi xati on in reading, Percept Psychophys 17:578-586, 1975. defects - a valuable strategy to improve reading ability and an indicati on for corti cal plasti city, Neuro - Ophthalmol 18:117-131, 20. McFadzean, R.M.: NovaVision: vision restorati on therapy, Current 1997. Opinion in Ophthalmology 17(6):498-503, 2006. 43. Trauzett el – Klosinski, S., Brendler, K.: Eye movements in reading with 21. Meienberg, O.: Augenbewegungsmuster bei Hemianopsie und hemianopic fi eld defects: the signifi cance of clinical parameters, visuellem Neglekt, Klin Monatsbl Augenheilk 192:108-112, 1988. Graefe`s Arch Clin Exp Ophthalmol 236: 91-102, 1998. 22. Nelles, G., Esser, J., Eckstein, A. et al. : Compensatory visual fi eld 44. Trauzett el – Klosinski, S., Reinhard, J.: The verti cal fi eld border in training for pati ents with hemianopia aft er stroke, Neurosci Lett human hemianopia and its signifi cance for fi xati on behavior and 306:189-192, 2001. reading, Invest Ophthalmol Vis Sci 39:2177-2186, 1998. 23. Pambakian, A.L.M., Wooding, D.S., Patel, N., et al.: Scanning the 45. Trauzett el-Klosinski, S.: Rehabilitati on in TBI-Hemianopia, North visual world: a study of pati ents with homonymous hemianopia, J American Neuro-Ophthalmology Society Annual Meeti ng Syllabus Neurol Neurosurg Psychiatry 69:751-759, 2000. 67-74, 2009. 24. Pambakian, A.L.M., Mannan, S.K., Hodgson, T.L., et al.: Saccadic 46. Trauzett el-Klosinski, S.: Rehabilitati on for Visual Disorders, J Neuro- visual search training: a treatment for pati ents with homonymous Ophthalmol 30:73-84, 2010. hemianopia, J Neurol Neurosurg Psychiatry 75:1443-1448, 2004. 47. Trobe, J.D., Lorber, M.L., Schlezinger, N.S.: Isolated Homonymous 25. Peli, E.: Field expansion for homonymous hemianopia by opti cally Hemianopia, Arch Ophthalmol 89 May:377-381, 1973. induced peripheral exotropia, Optometry and Vision Science 77:453, 2000. 48. Trobe, J.D., Sabel, B.A., Trauzett el-Klosinski, S.: Point counter point: Improving vision in a pati ent with homonymous hemianopia, J 26. Plant, G.T.: A work out for hemianopia, Br J Ophthalmol 89:2, 2005. Neuro-Ophthalmol 25:143-149, 2005. 27. Pilz, K., Braun, C., Altpeter, E. et al.: Modulati on of visual sti mulus 49. Vanni, S., Raninen, A., Näsänen, R. et al.: Dynamics of corti cal discriminati on by sustained focal att enti on: an MEG study, Invest acti vati on in a hemianopic pati ent, Neuro Report 12(4) Ophthalmol Vis Sci 47(3):1225-1229, 2006. March:861-865, 2001. 28. Pöppel, E., von Cramon, D., Backmund, H.: Eccentricity-specifi c 50. Weiskrantz, L.: Roots of blindsight, Progress in Brain Res 144:229- dissociati on of visual functi ons in pati ents with lesions of the central 241, 2004. visual pathways, Nature 256 August:489-490, 1975. 51. World Health Organisati on (WHO): Internati onal classifi cati on of 29. Raninen, A., Vanni, S., Hyvärinen, L. et al.: Temporal sensiti vity in a functi oning, disability and health (ICF), Geneva, 2004. hemianopic visual fi eld can be improved by long-term training using fl icker sti mulati on, J Neurol Neurosurg Psychiatry 78:66-73, 2007. 52. Zhang, X., Kedar, S., Lynn, M.J. et al.: Homonymous hemianopias, Clinical-anatomic correlati ons in 904 cases, Neurology Mar 28, 66(6): 30. Reinhard, J., Schreiber, A., Schiefer, U. et al.: Does visual resti tuti on 906-910, 2006. training change absolute homonymous scotoma? Briti sh Journal of Ophthalmology 89: 30-35, 2005. 53. Zihl, J., Krischer, C., Meißen, R.: Die hemianopische Lesestörung und ihre Behandlung, Nervenarzt 55:317-323, 1984. 31. Rossi, P.W., Kheyfets, S., Reding, M.J.: Fresnel prisms improve visual percepti on in stroke pati ents with homonymous hemianopia or 54. Zihl, J., von Cramon, D.: Resti tuti on of visual functi on in pati ents unilateral visual neglect, Neurology 40:1597, 1990. with cerebral blindness, J. Neurol. Neurosurg. Psychiatry 42:312-322, 1979. 32. Rosetti , Y., Rode, G., Pisella, L. et al.: Prism adaptati on to a rightward opti cal deviati on rehabilitates left hemispati al neglect, Nature 55. Zihl, J.: Visual scanning behaviour in pati ents with homonymous 395:166, 1998. hemianopia, Neuropsychologia 33: 287-303, 1995.

96 | North American Neuro-Ophthalmology Society CURRENT STATUS OF IMAGE-GUIDED RADIATION THERAPY FOR PATIENTS WITH BASE OF SKULL TUMORS Dean Cestari, MD The Massachusetts Eye and Ear Infirmary Boston, MA

LEARNING OBJECTIVES tumors in a more refi ned manner5. More recently, charged- parti cle-radiati on therapy is being used more commonly 1. The att endee will be able to learn the diff erences to treat many tumors traditi onally treated by ionizing between ionizing and charged parti cle (proton beam) radiati on6. radiati on. 2. The att endee will be able to learn the strategies used by radiati on oncologists to limit collateral damage to CLASSIFICATION normal structures when radiati ng a tumor. Various types of tumors involve the skull base and they 3. The att endee will be able to learn the advantages and include meningiomas, pituitary adenomas, paranasal sinus disadvantages of diff erent radiati on modaliti es. cancers, chordomas, chondrosarcomas, and schwannomas. There are many ways to classify and manage skull base tumors. Morita et al have suggested that these tumors can CME QUESTIONS be categorized according to their individual histopathologic types and locati ons1. Since there are numerous types of 1. What is the diff erence between ionizing and charged tumors, these authors have limited their classifi cati ons to parti cle radiati on? three groups based on biologic aggressiveness, and nine 2. What is Intensity-Modulated Photon Radiati on Therapy territories based on their locati on. They argue that biologic (IMRT)? aggressiveness directs our decision about which mode of treatment is the best from an oncological standpoint, and 3. What are the theoreti cal advantages of charged the locati on of tumor and clinical presentati on provide parti cle (proton beam) radiati on? informati on on the risks that are involved in treatment.

BIOLOGIC AGGRESSIVENESS KEYWORDS (MAX 5) Skull base tumors can be categorized into three pathologic 1. Skull Base enti ti es: benign, slow-growing (low-grade), and fast-growing (high-grade) malignancies1. 2. Ionizing Radiati on Benign tumors grow in an expansive fashion and induce 3. Charged Parti cle (Proton Beam) Radiati on. clinical symptoms by exerti ng pressure on various 4. Intensity-Modulated Photon Radiati on Therapy (IMRT) structures including the opti c nerves, chiasm and ocular motor nerves. Therefore, the reducti on of the mass eff ect and, if possible, the complete excision of the tumor with INTRODUCTION minimal morbidity are the gold standard of treatment. However, there are always associated surgical risks that Management of tumors in the cranial base has been may depend signifi cantly on the locati on and extent of the challenging for neuro-ophthalmologists, radiati on tumor. With the widespread use of magneti c resonance oncologists, otolaryngologists and neurosurgeons for more imaging (MRI), asymptomati c or minimally symptomati c than a century.1 Since tumors are located deep in the tumors such as meningiomas are more frequently being skull and can involve important visual and neurovascular discovered in the skull base. Treatment opti ons of these structures, surgical treatment was associated with high benign tumors include observati on, the excision of the morbidity and mortality. Visual pathways, both aff erent tumor by surgery, or the control of tumor growth by and eff erent, originate or traverse the cranial base and conformal radiotherapy. The long-term benefi ts and risks of are therefore aff ected by many of these tumors making each procedure and the natural history of various tumors management diffi cult. During the 1970s and 1980s, should be carefully assessed before choosing a specifi c neurosurgeons and head and neck surgeons2-4 introduced treatment opti on 1, 7-9. advances in surgical instrumentati on and developed new approaches to the cranial base. Simultaneously, focused stereotacti c radiotherapy was introduced to treat localized

2011 Annual Meeting Syllabus | 97 Slow-growing malignancies such as chondrosarcomas, and pulmonary embolus. CSF leakage complicates cranial chordomas, low-grade esthesioneuroblastomas, and base surgery in 5% to 30% of cases1, 10, whereas major adenoid cysti c carcinomas are best treated with a neurological complicati ons have occurred in 4% to 10% of combined mode of surgical debulking and radiotherapy10-12. cases1. Operati vely induced cranial neuropathies occur in Sensiti vity to radiotherapy is variable with each tumor 10% to 96% of cases, depending on the region of surgery, type1. Surgical resecti on should be as complete as possible, the nature of the tumor, and the aggressiveness of especially with radioresistant tumors such as chordomas. resecti on3. Many cases of adenoid cysti c carcinoma have extensive perineural invasion making complete resecti on unfeasible Due to the potenti al for major morbidity associated with without causing signifi cant morbidity. Radiotherapy for skull base surgery, an experienced treatment team should these tumors can involve either fracti onated or focused carefully determine the indicati on for and the extent of radiosurgery depending on the size and extent of the resecti on, and the benefi ts, risks, and ulti mate goals of tumor. Focused heavy parti cle irradiati on is reported the procedure should be thoroughly discussed with the with a favorable outcome in cases with chordomas, pati ent7, 8, 13. chondrosarcomas, and adenoid cysti c carcinoma 1, 13, 14. RADIOTHERAPY Highly malignant tumors should be removed en bloc if The goal of radiotherapy is to deliver a homogeneous the process does not involve criti cal structures. However, radiati on dose to a tumor target while minimizing the dose in most situati ons, piecemeal resecti on or sacrifi ce of a to surrounding normal ti ssue15. In this way, the maximum signifi cant structure may be involved in surgical resecti on. number of clonogenic tumor cells can be eradicated with Hence, radiotherapy with or without chemotherapy minimum risk to normal structures15. is the main treatment modality when the cranial base approach to a tumor may cause the loss of sensiti ve cranial EXTERNAL BEAM RADIATION THERAPY structures1. Ionizing radiati on uses either x-rays, gamma rays, or electron beams produced by a linear accelerator or MANAGEMENT OPTIONS OF CRANIAL BASE TUMORS radiati on sources such as radioacti ve seeds to treat tumors. For pati ents presenti ng with skull base tumors, treatment Ionizing radiati on preferenti ally damages the DNA of tumor opti ons include observati on with close follow-up, operati ve cells that ulti mately causes cell death. The radiati on dose is excision, radiotherapy, chemotherapy, or a combinati on of measured in Gray (Gy) and cellular damage increases with these therapies1. higher absorbed radiati on doses.16 When coming up with a treatment plan, radiati on oncologists must balance the OBSERVATION desired damage to the tumor and the undesirable damage For those skull base tumors that are suspected to be to adjacent healthy ti ssue 17. This is typically accomplished benign, observati on is a reasonable opti on, parti cularly by targeti ng the beam to the tumor area using paths that when the pati ent has minimal symptoms. Skull base spare nearby criti cal and radiosensiti ve structures, selecti ng meningiomas and schwannomas are the most typical multi ple fi elds that cross in the tumor area through examples. Careful neuro-ophthalmic follow-up, including diff erent paths, and splitti ng the total dose into smaller visual fi elds and imaging studies, is important in detecti ng dose “fracti ons” that are delivered over several days or changes and in following the progress of the tumor1, 9. weeks16.

SURGERY THREE-DIMENSIONAL CONFORMAL RADIATION THERAPY Since cranial base tumors are deep in locati on and (3DCRT) surrounded by criti cal structures, conventi onal Conformal radiotherapy att empts to minimize the neurosurgical approaches are disadvantageous because volume of normal ti ssue irradiated, by shaping the dose of the need for signifi cant brain retracti on, poor control of distributi on to conform ti ghtly to the shape of the tumor, the lesion and adjacent structures, and oft en subopti mal reducing the dose to surrounding normal ti ssues18. exposure. Modern skull base surgery carries certain risks. Adequate immobilizati on of the target and improved Because the approaches used in these surgeries require three-dimensional (3D) imaging enable a higher degree of extensive bone removal and oft en necessitate entrance certainty of target localizati on, which permits the use of into or transgression of potenti ally contaminated spaces narrower margins around the target19. Using informati on (e.g., the paranasal sinuses), the potenti al exists for from CT, MRI or PET scans, 3DCRT is a technique where complicati ons such as cerebrospinal fl uid (CSF) leakage, the beams of radiati on used in treatment are shaped to infecti on, cosmeti c defects, and cranio-vertebral instability. match the tumor. 3D dose distributi ons are calculated by a The exposure or manipulati on of neurovascular structures treatment planning computer using dosimetric algorithms. also carries risks for stroke or cranial neuropathies. The Previously, radiati on treatment matched the height and operati ve ti me and extent are increased and pose risks width of the tumor, meaning that healthy ti ssue had to be for perioperati ve medical problems such as pneumonia, exposed to the beams. Conformal radiati on therapy uses hormonal and electrolyte disturbance, coagulopathy, the targeti ng informati on to focus precisely on the tumor

98 | North American Neuro-Ophthalmology Society while minimizing damage to the healthy surrounding ti ssue. Bragg who discovered it in 1903, plots the energy loss of This exact targeti ng makes it possible to use higher levels ionizing radiati on during its travel through matt er (Figure of radiati on in treatment while minimizing injury to healthy 1). For protons, α-rays, and other ion rays, the peak occurs or sensiti ve structures. Radiotherapy planning studies have immediately before the parti cles come to rest. By adjusti ng confi rmed that 3DCRT reduces the volume of normal ti ssue the energy of the charged parti cles and the intensity of within the high dose volume compared with conventi onal the beam, one can deliver specifi ed doses anywhere in the radiotherapy20-22. However, it is impossible to spare normal body with high precision16, 23. ti ssue in close proximity to the tumor target because the beams cannot be modulated with this technique. Charged parti cles damage cell DNA in qualitati vely diff erent ways than photons or electrons. Therefore the same INTENSITY-MODULATED PHOTON RADIATION THERAPY amount of physical radiati on can cause greater cellular (IMRT) damage. The relati ve biological eff ecti veness (RBE) is the Historically, the maximum radiati on dose that could be rati o of the dose required to produce a specifi c biological given to a tumor site has been restricted by the tolerance eff ect. Photons are used as the reference radiati on and the and sensiti vity of the surrounding nearby healthy ti ssues. RBE is 1.1 meaning protons result in approximately 10% IMRT is a form of conformal therapy that combines several more biological damage per unit dose than photons24. intensity modulated beams15. With IMRT, the radiati on beam can be broken up into many “beamlets,” and the Charged parti cle radiati on therapy is expected to deliver intensity of each beamlet can be adjusted individually. biologically equivalent doses more precisely and with less With this technique, conformal photon radiati on is damage to normal structures than conventi onal photon delivered to the target tumor by crossing multi ple properly radiati on therapy14, 16, 25. It is believed that its ability to shaped radiati on fi elds with modulated intensiti es through maximize the dose and target the tumor with high precision paths that spare radiosensiti ve and criti cal adjoining ti ssue. allows for bett er tumor control. Theoreti cally, this could The intensity of the radiati on in IMRT can be changed be benefi cial in children because they are considered during treatment to spare more adjoining normal ti ssue more suscepti ble to radiati on side eff ects including the than is spared during conventi onal radiati on therapy. By development of secondary malignancies26. It is unclear shaping the radiati on beams to closely approximate the whether the claimed high precision in dose delivery is shape of the tumor, an increased dose of radiati on can benefi cial for all tumors in adults and several investi gators be delivered to the tumor with less damage to normal have suggested that proton-beam radiati on therapy may structures16. IMRT therefore off ers a signifi cant advance in be indicated in approximately 15% of pati ents undergoing conformal therapy by improving conformality and reducing irradiati on27. radiati on dose to radiosensiti ve normal structures close close to the tumor15. Charged-parti cle radiati on therapy is an alternati ve mode of radiati on therapy that is becoming increasingly available. CHARGED PARTICLE RADIATION THERAPY Seven proton-beam faciliti es are in operati on in the United Charged-parti cle-radiati on therapy includes external States as of July 2009 and at least 4 are under constructi on. radiotherapy that uses protons, helium, carbon, neon, The theoreti cal advantage of this type of radiati on silicon ions or other charged parti cles. Currently, only therapy over traditi onal photon treatment has yet to be protons and carbon ions are in clinical use today and demonstrated in clinical trials but several are currently this charged-parti cle-beam therapy has been clinically ongoing and the results should be available soon. available since 195416 but was not approved as a radiati on treatment opti on by the FDA unti l 1988. Charged parti cles represent an advancement over photons because they have a superior depth-dose distributi on. Photons (or electron beams) deposit most of their energy near the surface with progressively smaller doses at larger depths and they conti nue to deposit the dose of radiati on in normal ti ssues beyond the tumor. Charged parti cles deposit a low dose near the surface and deposit almost all of their energy in the fi nal millimeters of their trajectory in the tumor. There is essenti ally no exit dose and ti ssues beyond the tumor receive very litt le of the dose. This patt ern results in a sharp and localized peak dose, known as the Bragg peak. When a fast charged parti cle moves through matt er, it ionizes atoms of the material and deposits a Figure 1: Diagram of the energy deposit of electrons, photons and dose along its path. A peak occurs because the interacti on protons. Almost the enti re energy of protons is liberated in a very cross secti on increases as the charged parti cle’s energy narrow peak known as the Bragg peak. decreases. The Bragg peak, named aft er William Henry

2011 Annual Meeting Syllabus | 99 CME ANSWERS 10. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: results and follow-up of 60 1. Ionizing radiati on uses either x-rays, gamma rays or pati ents. Neurosurgery. May 1995;36(5):887-896; discussion 896- electron beams produced by a linear accelerator 887. whereas charged-parti cle-radiati on can consist of 11. Gormley WB, Sekhar LN, Wright DC, et al. Management and long- protons, helium, carbon, neon, or silicon ions or other term outcome of adenoid cysti c carcinoma with intracranial extension: a neurosurgical perspecti ve. Neurosurgery. Jun charged parti cles. Currently, only protons and carbon 1996;38(6):1105-1112; discussion 1112-1103. ions are in clinical use today. 12. Morita A, Ebersold MJ, Olsen KD, Foote RL, Lewis JE, Quast LM. Esthesioneuroblastoma: prognosis and management. Neurosurgery. 2. Three dimensional (3-D) conformal radiati on therapy May 1993;32(5):706-714; discussion 714-705. is a technique where the beams of radiati on used 13. Cusimano MD, Sekhar LN, Sen CN, et al. The results of surgery in treatment are shaped to match the tumor while for benign tumors of the cavernous sinus. Neurosurgery. Jul minimizing damage to the healthy surrounding ti ssue. 1995;37(1):1-9; discussion 9-10. With IMRT, the radiati on beam can be broken up into 14. Pommier P, Liebsch NJ, Deschler DG, et al. Proton beam radiati on many “beamlets,” and the intensity of each beamlet therapy for skull base adenoid cysti c carcinoma. Arch Otolaryngol can be adjusted individually thereby limiti ng collateral Head Neck Surg. Nov 2006;132(11):1242-1249. damage to nearby normal structures. 15. Nutti ng C, Dearnaley DP, Webb S. Intensity modulated radiati on therapy: a clinical review. Br J Radiol. May 2000;73(869):459-469. 3. Charged parti cles represent an advancement over 16. Terasawa T, Dvorak T, Ip S, Raman G, Lau J, Trikalinos TA. Systemati c photons because they have a superior depth-dose review: charged-parti cle radiati on therapy for cancer. Ann Intern Med. Oct 20 2009;151(8):556-565. distributi on. Photons (or electron beams) deposit most 17. Levin WP, Kooy H, Loeffl er JS, DeLaney TF. Proton beam therapy. Br J of their energy near the surface with progressively Cancer. Oct 17 2005;93(8):849-854. smaller doses at larger depths and they conti nue to 18. Purdy JA. Advances in three-dimensional treatment planning and deposit the dose of radiati on in normal ti ssues beyond conformal dose delivery. Semin Oncol. Dec 1997;24(6):655-671. the tumor. Charged parti cles deposit a low dose near 19. Horwitz EM, Hanlon AL, Hanks GE. Update on the treatment of the surface and deposit almost all of their energy in prostate cancer with external beam irradiati on. Prostate. Nov 1 the fi nal millimeters of their trajectory in the tumor. 1998;37(3):195-206. Tissues beyond the tumor receive very litt le of the 20. Dearnaley DP. Radiotherapy of prostate cancer: established results dose. and new developments. Semin Surg Oncol. Jan-Feb 1995;11(1):50- 59. 21. Emami B, Purdy JA, Simpson JR, Harms W, Gerber R, Wippold JF. 3-D REFERENCES conformal radiotherapy in head and neck cancer. The Washington University experience. Front Radiat Ther Oncol. 1996;29:207-220. 1. Morita A, Sekhar LN, Wright DC. Current Concepts in the Management of Tumors of the Skull Base. Cancer Control. Mar 22. Lawrence TS, Kessler ML, Robertson JM. 3-D conformal radiati on 1998;5(2):138-149. therapy in upper gastrointesti nal cancer. The University of Michigan experience. Front Radiat Ther Oncol. 1996;29:221-228. 2. Fisch U. Infratemporal fossa approach to tumours of the temporal bone and base of the skull. J Laryngol Otol. Nov 1978;92(11):949- 23. Schulz-Ertner D, Nikoghosyan A, Thilmann C, et al. Results of carbon 967. ion radiotherapy in 152 pati ents. Int J Radiat Oncol Biol Phys. Feb 1 2004;58(2):631-640. 3. Hakuba A, Nishimura S, Tanaka K, Kishi H, Nakamura T. Clivus meningioma: six cases of total removal. Neurol Med Chir (Tokyo). 24. Schulz-Ertner D, Tsujii H. Parti cle radiati on therapy using proton and 1977;17(1 Pt 1):63-77. heavier ion beams. J Clin Oncol. Mar 10 2007;25(8):953-964. 4. Sekhar LN, Swamy NK, Jaiswal V, Rubinstein E, Hirsch WE, Jr., 25. Suit H, Kooy H, Trofi mov A, et al. Should positi ve phase III clinical Wright DC. Surgical excision of meningiomas involving the trial data be required before proton beam therapy is more widely clivus: preoperati ve and intraoperati ve features as predictors adopted? No. Radiother Oncol. Feb 2008;86(2):148-153. of postoperati ve functi onal deteriorati on. J Neurosurg. Dec 26. Geenen MM, Cardous-Ubbink MC, Kremer LC, et al. Medical 1994;81(6):860-868. assessment of adverse health outcomes in long-term survivors of 5. Lunsford LD. Contemporary management of meningiomas: radiati on childhood cancer. JAMA. Jun 27 2007;297(24):2705-2715. therapy as an adjuvant and radiosurgery as an alternati ve to 27. Glimelius B, Ask A, Bjelkengren G, et al. Number of pati ents surgical removal? J Neurosurg. Feb 1994;80(2):187-190. potenti ally eligible for proton therapy. Acta Oncol. 2005;44(8):836- 6. Chan AW, Liebsch NJ. Proton radiati on therapy for head and neck 849. cancer. J Surg Oncol. Jun 15 2008;97(8):697-700. 7. Ojemann RG. Skull-base surgery: a perspecti ve. J Neurosurg. Apr 1992;76(4):569-570. 8. Ojemann RG, Black PM. Diffi cult decisions in managing pati ents with benign brain tumors. Clin Neurosurg. 1989;35:254-284. 9. Miller NR. New concepts in the diagnosis and management of opti c nerve sheath meningioma. J Neuroophthalmol. Sep 2006;26(3):200- 208.

100 | North American Neuro-Ophthalmology Society STEM CELLS, GENE THERAPY AND THE NEURO-OPHTHALMOLOGIST “I fi nd nothing more depressing than opti mism” —Paul Fussell

Jeffrey Bennett, MD, PhD University of Colorado Denver Aurora, CO

LEARNING OBJECTIVES KEYWORDS 1. The att endee will be able to enumerate the challenges 1. Opti c Neuropathy facing functi onal opti c nerve restorati on. 2. Axon 2. The att endee will be able to elaborate strategies for stem cell therapy for opti c nerve disorders. 3. Stem Cells 3. The att endee will be able to list potenti al gene therapy 4. Gene Therapy targets for maintenance of reti nal ganglion cell survival 5. Reti nal Ganglion Cells and axonal outgrowth following opti c neuropathy.

Neural regenerati on and repair has remained the “holy CME QUESTIONS grail” of neuro-ophthalmic therapy. Recent advances in 1. Therapeuti c hurdles facing opti c nerve restorati on our understanding of neuronal development and survival, include all of the following except: axonal guidance, disease pathogenesis, and medical technology are allowing us to view opti c nerve restorati on a. Functi onal innervati on of target ti ssue as something more than a therapeuti c oxymoron. b. Modifi cati on of the ti ssue environment in the injured opti c nerve Opti c nerve injuries are a signifi cant cause of visual morbidity. Using ocular coherence tomography c. Promoti ng reti nal ganglion cells axon extension (OCT), researchers have demonstrated that ischemic, infl ammatory, toxic, and degenerati ve opti c neuropathies d. Accentuati ng astrocyte hypertrophy result in signifi cant reti nal nerve fi ber layer (RNFL) loss that correlates with loss of visual functi on.1-9 In individuals with 2. A second messenger important in multi ple steps of opti c neuriti s or multi ple sclerosis, signifi cant RNFL loss may opti c nerve regenerati on is occur in the absence of diminished high contrast acuity2, 7 a. Calcium indicati ng that eff ecti ve restorati ve therapy for opti c nerve injury is important for acute and chronic opti c nerve injuries b. Phosphati dylinositol independent of the fi nal visual outcome. What is the status c. Cyclic AMP of current eff orts to bring restorati ve therapy to pati ents with opti c nerve injury? And what are the challenges facing d. Tyrosine phosphorylati on future progress in the fi eld? In this review, we will examine the knowledge, tools, and techniques that are currently 3. Potenti al sources of adult stem cells include all of the available to restore functi on to the damaged opti c nerve following except: and examine the challenges that remain for future research. a. Dentate gyrus OPTIC NERVE RESTORATION: A COMPLEX PROBLEM b. Reti nal pigment epithelium Functi onal restorati on or regenerati on of opti c nerve ti ssue following injury is a complex problem that requires the c. Fibroblasts considerati on of multi ple interdependent factors. Any d. Bone marrow successful strategy necessitates the eff ecti ve modulati on of multi ple pathways involved in diverse processes such as injury response, infl ammati on, cell death, neural development, and axonal pathfi nding (Figure).10, 11

2011 Annual Meeting Syllabus | 101 overexpression of Bcl-2,13 an anti apoptoti c protein, or delivery of neurotrophic factors such as brain-derived neurotrophic factor (BDNF),15 neurotrophin 4/5 (NT-4/5),16 nerve growth factor (NGF), ciliary neurotrophic factor (CNTF)16 and glial-derived neurotrophic factor (GDNF)17 have shown parti al success in reducing RGC loss following axotomy in animal models. Unfortunately, technical, strategic and physiologic hurdles need to be overcome before human clinical trials can be developed. Successful gene therapy will require novel strategies to control the distributi on and ti ming of gene expression, as well as new technologies to deliver suffi cient transgenes to specifi c target cells. Successful neurotrophic therapy will require concurrent treatment to enhance the ability of injured RGCs to respond to trophic support. Following injury, RGCs downregulate their neurotrophin (Trk) receptors,18 and exogenous administrati on of neurotrophic factors may result in additi onal receptor tachyphylaxis.19, 20 Figure. Functi onal opti c nerve restorati on: challenges and These phenomena may be counteracted by recruiti ng potenti al interventi ons. Trk receptors to the cell surface through elevati on of intracellular cyclic-AMP (cAMP)21 or axonal depolarizati on.22 A combined approach of neurotrophic factor administrati on,23 pharmacologic elevati on of cAMP,24 and First, the injurious mechanism must be alleviated and transcorneal electrical sti mulati on25, 26 may be required. To interventi ons initi ated to opti mize reti nal ganglion date, the administrati on of neurotrophic factors to human cell (RGC) survival. The ti ming and nature of these pati ents has been limited to small studies focused on reti nal interventi ons may diff er based on both the mode and degenerati ve disorders. locati on of injury to the RGC.12 Second, damaged ti ssue needs to be modifi ed to maximize RGC regenerati on, stem CELL REPLACEMENT THERAPY cell diff erenti ati on, and axonal growth. Third, surviving In conjuncti on with RGC preservati on, successful RGCs or transplanted stem cells need to be sti mulated regenerati ve therapy will likely require additi onal to extend axonal processes. And fourth, these processes interventi ons to restore RGC numbers. Cellular require correct guidance towards appropriate targets in replacement therapy using multi potent progenitor cells the lateral geniculate nucleus and superior colliculus. The or stem cells has emerged as the primary method for complex orchestrati on of events needed for the restorati on achieving this goal. Multi potent progenitor cells can of opti c nerve functi on following injury is the main reason diff erenti ate into many cell types, while stem cells are why recent eff orts have met with minimal success. For a self-renewing populati on that are pluripotent. These instance, to date, experimental interventi ons have been cells might promote functi onal recovery by reconsti tuti ng directed at a limited number of key hurdles such as RGC depleted RGCs, remyelinati ng axons, or facilitati ng axon survival13 or ti ssue replacement14. regenerati on. Successful cell replacement therapy requires the survival, integrati on and diff erenti ati on of transplanted RETINAL GANGLION CELL SURVIVAL cells. The diff erenti ati on of transplanted embryonic Opti c neuropathies may result from direct or indirect RGC stem cells is parti cularly criti cal since undiff erenti ated injury at the level of the cell body or axon. In trauma, RGC embryonic stem cells are potenti ally teratogenic.27 To axons are injured directly by mechanical stress; while in limit this potenti al, embyronic stem cells are typically pre- neuromyeliti s opti ca, RGC axons are injured indirectly due diff erenti ated to a desired lineage before transplantati on. to the loss of trophic astrocyti c support. In glaucoma, Several groups have reported the appropriate however, mechanical stress, vascular insuffi ciency, oxidati ve diff erenti ati on of embryonic stem cells28, 29 or diff erenti ated injury, and infl ammati on may combine to damage RGCs progenitor cells30 following transplantati on into injured at both the level of the cell body and axon. Following rodent spinal cords. Transplantati on of diff erenti ated injury, cell death may proceed through one of several embryonic stem cells into the reti na31 and axotomized opti c disti nct pathways: programmed cell death (apoptosis), tract32 has been shown to promote photoreceptor and RGC necrosis, or secondary degenerati on.12 Initi al therapeuti c survival in rodent models; however, no engraft ment of the strategies have focused on inhibiti ng apoptosis due to the diff erenti ated stem cells was observed. major role of programmed cell death in determining RGC fate following opti c nerve injury. Strategies have included Adult stem cells may provide an additi onal source of direct modulati on of apoptoti c cell pathway signaling and material for cell transplantati on. In contrast to embryonic delivery of trophic factors to injured RGCs. For example, stem cells, adult stem cells may be recovered from any

102 | North American Neuro-Ophthalmology Society individual and used for autologous transplantati on. Sources additi on to NgR.51 A second explanati on is the presence of may include reti na,33, 34 adult central nervous system,35 and additi onal redundant inhibitory signaling pathways. For skin.36 Intraocular-administered adult hippocampal stem instance, inhibiti on of epidermal growth factor receptor cells were reported to engraft in the reti nal ganglion cell and protein kinase C signaling have been shown to benefi t layer following opti c nerve axotomy or superior colliculus axonal outgrowth aft er opti c nerve and spinal cord injury.52, ablati on.37 In rodent and non-human primate models 53 since most of these signaling pathways converge on the of infl ammatory central nervous system demyelinati on small GTPase, Rho, therapeuti c inhibiti on of Rho signaling (experimental autoimmune encephalomyeliti s), neural stem is a focus of acti ve investi gati on.54 The bacterial toxin cells ameliorated disease and promoted neuronal survival C3-ADP-ribosyltransferase55 and cAMP elevati on56 have through immunomodulatory eff ects.38, 39 demonstrated some promise.

Adult bone marrow contains several diff erent stem Tissue graft s off er an opportunity to bypass the inhibitory cell populati ons: hematopoieti c stem cells (HSCs) and environment of the injured opti c nerve. In seminal mesenchymal stem cells (MSCs). MSCs are multi potenti al experiments, Aguayo et al 57, 58 demonstrated that a stromal precursor cells that are common precursors of signifi cant number of RGCs could regrow axons through multi ple mesenchymal ti ssues such as fat, bone, carti lage peripheral nerve graft s to reinnervate targets following and connecti ve ti ssues. MSCs have been subsequently axotomy. Peripheral nerve ti ssue provides a permissive reported in many ti ssues, including the vasculature.40 In substrate for axonal growth due to the salutatory nature additi on, MSCs may have the ability to diff erenti ate into of the local ti ssue environment and the trophic support of other germ lineages such as neural cells,41, 42 although the Schwann cells. Schwann cells may also act to change the experimental results are not defi niti ve.43 The potenti al injury response of astrocytes and inhibit the formati on of mechanisms by which HSCs and MSCs act are sti ll unclear, glial scar.59 Potenti al alternati ve substrates to peripheral although the elaborati on of factors that facilitate neuronal nerve ti ssue include perinatal opti c nerve,60 geneti cally- and glial survival, axon elongati on, and immunoregulati on engineered Schwann cells,61 fetal brain graft s,62 olfactory has been proposed as the primary mechanism.44 ensheathing cells,63 and Schwann cell impregnated In preclinical trials in multi ple sclerosis, MSCs have matrices.64 demonstrated benefi cial immunoregulatory properti es with limited cell replacement.45 Nevertheless, MSCs may prove REPROGRAMMING THE ADULT RETINAL GANGLION CELL to be a powerful tool to limit RGC loss and promote axonal A major impediment to the successful restorati on of opti c regenerati on aft er opti c neuriti s. nerve ti ssue is the inability of adult RGCs to extend new axons. Embryonic reti nal explants65 or purifi ed embryonic MODIFYING THE INJURED OPTIC NERVE RGCs66 extend axons more effi ciently than their adult In contrast to the permissive environment of the peripheral counterparts independent of the extrinsic environment. nervous system (PNS), the mature central nervous system The inability of adult RGCs to extend new axons arises from (CNS) is not conducive to axonal regenerati on. Following an intrinsic developmental program. Gene therapy may opti c nerve injury, inhibitory glial proteins such as myelin- off er one avenue for reprogramming adult RGCs to behave associated glycoprotein (MAG), Nogo-A, semaphorins, and more like their embryonic counterparts. Recently, Moore oligodendrocyte myelin glycoprotein (Omgp) are released et al 67 identi fi ed two members of Krüppel-like family of at the site of injury and proliferati ng astrocytes secrete transcripti on factors, KLF4 and KLF9, that decreased neurite inhibitory extracellular matrix proteoglycans.46 In additi on, outgrowth when overexpressed in postnatal RGCs. In acti vated glial cells may express axonal guidance proteins additi on, gene deleti on of KLF4 from RGCs increased axonal in a manner that suppresses regenerati on. For instance, regenerati on following injury. Downregulati on of KLF4 semaphorin-3A and netrin may inhibit RGC restorati on or KLF9 in injured in RGCs by siRNA inhibiti on may off er by acti ng as repulsive cues to regenerati ng axons. MAG, a novel approach for promoti ng axonal growth following Nogo-A and Omgp inhibit axonal outgrowth through opti c neuropathy. acti vati on of the Nogo receptor (NgR). Experimental inhibiti on of MAG and Nogo signaling, however, has had An alternati ve approach to enhancing the regenerati ve variable success depending on the mode of inhibiti on. capacity of adult RGCs is to upregulate or downregulate the Chromophore-assisted laser inacti vati on of MAG in expression of developmental genes. Modulati ng the level embryonic chick explants increased axonal regenerati on of cAMP in injured RGCs may simultaneously upregulate following opti c nerve crush injury;47 however, opti c nerve neurotrophin receptors and reduce responsiveness to regenerati on following crush injury was not facilitated in MAG/Nogo-A inhibiti on through upregulati on of arginase MAG knockout mice.48 Similarly, axonal regenerati on in I expression.68 Acti vati on of MAP kinase signaling or animals with geneti c deleti on of either Nogo-A or NgR mammalian target of rapamycin (mTOR) signaling may also was inferior to anti body-mediated neutralizati on.49, 50 One promote axonal regenerati on and RGC survival.69, 70 explanati on may lie in the complexiti es of the MAG and Nogo-A signaling networks. For instance, MAG may signal through sialoglycans (gangliosides GD1a and GT1b) in

2011 Annual Meeting Syllabus | 103 DIRECTING REGENERATING OPTIC NERVE AXONS TO 7. Pulicken M, Gordon-Lipkin E, Balcer LJ, et al. Opti cal coherence APPROPRIATE TARGETS tomography and disease subtype in multi ple sclerosis. Neurology 2007;69:2085-2092. Functi onal opti c nerve restorati on requires the reestablishment of appropriate axonal innervati on of target 8. Sadun AA, Salomao SR, Berezovsky A, et al. Subclinical carriers and conversions in Leber hereditary opti c neuropathy: a prospecti ve neurons in the lateral geniculus and superior colliculus. psychophysical study. Transacti ons of the American Ophthalmological Regenerati ng axons must exit the globe at the opti c Society 2006;104:51-61. nerve head and be guided through the opti c chiasm in a 9. Zoumalan CI, Agarwal M, Sadun AA. Opti cal coherence tomography topographic patt ern to their ipsilateral and contralateral can measure axonal loss in pati ents with ethambutol-induced opti c targets. Molecules important for various steps in this neuropathy. Graefes Arch Clin Exp Ophthalmol 2005;243:410-416. process have been identi fi ed: netrin-1 at the opti c nerve 10. Bull ND, Marti n KR. Opti c nerve restorati on: new perspecti ves. J head,71 semiphorin-5A and Slit in the opti c nerve,72, 73 and Glaucoma 2007;16:506-511. ephrin-B2 at the opti c chiasm,74 and ephrin-A ligands in the 11. Moore DL, Goldberg JL. Four steps to opti c nerve regenerati on. J superior colliculus.75 In adult hamsters, regenerati ng axons Neuroophthalmol 2010 extending through peripheral nerve graft s were noted 12. Whitmore AV, Libby RT, John SWM. Glaucoma: thinking in new to migrate to target ti ssue and form functi onal synapses. ways-a rôle for autonomous axonal self-destructi on and other While the percentage of transiti ng axons were low and the compartmentalised processes? Prog Reti n Eye Res 2005;24:639-662. axonal arbors limited, successful reinnervati on suggests 13. Bonfanti L, Strett oi E, Chierzi S, et al. Protecti on of reti nal ganglion cells from natural and axotomy-induced cell death in neonatal that appropriate cues are sti ll available to direct axonal transgenic mice overexpressing bcl-2. J Neurosci 1996;16:4186-4194. guidance should axonal regenerati on succeed. 14. Berry M, Carlile J, Hunter A. Peripheral nerve explants graft ed into the vitreous body of the eye promote the regenerati on of In conclusion, functi onal opti c nerve restorati on is no reti nal ganglion cell axons severed in the opti c nerve. J Neurocytol longer the subject of therapeuti c fantasy. To succeed, 1996;25:147-170. signifi cant yet not insurmountable hurdles must be 15. Nakazawa T, Tamai M, Mori N. Brain-derived neurotrophic factor overcome. Advancements in our understanding of neuro- prevents axotomized reti nal ganglion cell death through MAPK and ophthalmologic disease will contribute signifi cantly to PI3K signaling pathways. Invest Ophthalmol Vis Sci 2002;43:3319- 3326. therapies aimed at limiti ng RGC loss. Technological progress in gene therapy and cell replacement technology 16. Mey J, Thanos S. Intravitreal injecti ons of neurotrophic factors support the survival of axotomized reti nal ganglion cells in adult rats will soon lead to human clinical trials with signifi cant hope in vivo. Brain Res 1993;602:304-317. for functi onal improvement aft er opti c neuropathy. 17. Klöcker N, Bräunling F, Isenmann S, Bähr M. In vivo neurotrophic eff ects of GDNF on axotomized reti nal ganglion cells. Neuroreport CME ANSWERS 1997;8:3439-3442. 1) d 18. Cui Q, Tang LS, Hu B, et al. Expression of trkA, trkB, and trkC in injured and regenerati ng reti nal ganglion cells of adult rats. Invest 2) c Ophthalmol Vis Sci 2002;43:1954-1964. 19. Pernet V, Di Polo A. Synergisti c acti on of brain-derived neurotrophic 3) b factor and lens injury promotes reti nal ganglion cell survival, but leads to opti c nerve dystrophy in vivo. Brain : a journal of neurology 2006;129:1014-1026. REFERENCES 20. Spalding KL, Cui Q, Harvey AR. Reti nal ganglion cell neurotrophin receptor levels and trophic requirements following target ablati on in 1. Barboni P, Savini G, Valenti no ML, et al. Reti nal nerve fi ber layer the neonatal rat. Neuroscience 2005;131:387-395. evaluati on by opti cal coherence tomography in Leber’s hereditary opti c neuropathy. Ophthalmology 2005;112:120-126. 21. Meyer-Franke A, Kaplan MR, Pfrieger FW, Barres BA. Characterizati on of the signaling interacti ons that promote the survival and growth of 2. Costello F, Coupland S, Hodge W, et al. Quanti fying axonal loss developing reti nal ganglion cells in culture. Neuron 1995;15:805-819. aft er opti c neuriti s with opti cal coherence tomography. Ann Neurol 2006;59:963-969. 22. Shen S, Wiemelt AP, McMorris FA, Barres BA. Reti nal ganglion cells lose trophic responsiveness aft er axotomy. Neuron 1999;23:285-295. 3. Deleón-Ortega J, Carroll KE, Arthur SN, Girkin CA. Correlati ons between reti nal nerve fi ber layer and visual fi eld in eyes with 23. Tao W. Applicati on of encapsulated cell technology for reti nal nonarteriti c anterior ischemic opti c neuropathy. Am J Ophthalmol degenerati ve diseases. Expert Opin Biol Ther 2006;6:717-726. 2007;143:288-294. 24. Monsul NT, Geisendorfer AR, Han PJ, et al. Intraocular injecti on of 4. Hood DC, Anderson S, Rouleau J, et al. Reti nal nerve fi ber structure dibutyryl cyclic AMP promotes axon regenerati on in rat opti c nerve. versus visual fi eld functi on in pati ents with ischemic opti c Exp Neurol 2004;186:124-133. neuropathy. A test of a linear model. Ophthalmology 2008;115:904- 25. Miyake K-i, Yoshida M, Inoue Y, Hata Y. Neuroprotecti ve eff ect of 910. transcorneal electrical sti mulati on on the acute phase of opti c nerve 5. Horowitz J, Fishelzon-Arev T, Rath EZ, et al. Comparison of opti c injury. Invest Ophthalmol Vis Sci 2007;48:2356-2361. nerve head topography fi ndings in eyes with non-arteriti c anterior 26. Tagami Y, Kurimoto T, Miyoshi T, et al. Axonal regenerati on induced ischemic opti c neuropathy and eyes with glaucoma. Graefes Arch Clin by repeti ti ve electrical sti mulati on of crushed opti c nerve in adult Exp Ophthalmol 2010;248:845-851. rats. Jpn J Ophthalmol 2009;53:257-266. 6. Milea D, Sander B, Wegener M, et al. Axonal loss occurs early in dominant opti c atrophy. Acta Ophthalmol 2010;88:342-346.

104 | North American Neuro-Ophthalmology Society 27. Arnhold S, Klein H, Semkova I, et al. Neurally selected embryonic 48. Bartsch U, Bandtlow CE, Schnell L, et al. Lack of evidence that myelin- stem cells induce tumor formati on aft er long-term survival following associated glycoprotein is a major inhibitor of axonal regenerati on in engraft ment into the subreti nal space. Invest Ophthalmol Vis Sci the CNS. Neuron 1995;15:1375-1381. 2004;45:4251-4255. 49. Teng FYH, Tang BL. Why do Nogo/Nogo-66 receptor gene knockouts 28. McDonald JW, Liu XZ, Qu Y, et al. Transplanted embryonic stem cells result in inferior regenerati on compared to treatment with survive, diff erenti ate and promote recovery in injured rat spinal cord. neutralizing agents? J Neurochem 2005;94:865-874. Nat Med 1999;5:1410-1412. 50. Buchli AD, Schwab ME. Inhibiti on of Nogo: a key strategy to increase 29. Harper JM, Krishnan C, Darman JS, et al. Axonal growth of embryonic regenerati on, plasti city and functi onal recovery of the lesioned stem cell-derived motoneurons in vitro and in motoneuron-injured central nervous system. Ann Med 2005;37:556-567. adult rats. Proc Natl Acad Sci USA 2004;101:7123-7128. 51. Schnaar RL, Lopez PHH. Myelin-associated glycoprotein and its axonal 30. Cao Q, Benton RL, Whitt emore SR. Stem cell repair of central nervous receptors. J Neurosci Res 2009;87:3267-3276. system injury. J Neurosci Res 2002;68:501-510. 52. Sivasankaran R, Pei J, Wang KC, et al. PKC mediates inhibitory 31. Meyer JS, Katz ML, Maruniak JA, Kirk MD. Embryonic stem cell- eff ects of myelin and chondroiti n sulfate proteoglycans on axonal derived neural progenitors incorporate into degenerati ng reti na and regenerati on. Nat Neurosci 2004;7:261-268. enhance survival of host photoreceptors. Stem Cells 2006;24:274- 53. Koprivica V, Cho K-S, Park JB, et al. EGFR acti vati on mediates 283. inhibiti on of axon regenerati on by myelin and chondroiti n sulfate 32. Hill AJ, Zwart I, Samaranayake AN, et al. Rat neurosphere cells protect proteoglycans. Science 2005;310:106-110. axotomized rat reti nal ganglion cells and facilitate their regenerati on. 54. Dergham P, Ellezam B, Essagian C, et al. Rho signaling pathway J Neurotrauma 2009;26:1147-1156. targeted to promote spinal cord repair. J Neurosci 2002;22:6570- 33. Limb GA, Salt TE, Munro PMG, et al. In vitro characterizati on of a 6577. spontaneously immortalized human Müller cell line (MIO-M1). Invest 55. Winton MJ, Dubreuil CI, Lasko D, et al. Characterizati on of new cell Ophthalmol Vis Sci 2002;43:864-869. permeable C3-like proteins that inacti vate Rho and sti mulate neurite 34. Mayer EJ, Carter DA, Ren Y, et al. Neural progenitor cells outgrowth on inhibitory substrates. J Biol Chem 2002;277:32820- from postmortem adult human reti na. The Briti sh journal of 32829. ophthalmology 2005;89:102-106. 56. Hu Y, Cui Q, Harvey AR. Interacti ve eff ects of C3, cyclic AMP and 35. Pluchino S, Quatt rini A, Brambilla E, et al. Injecti on of adult ciliary neurotrophic factor on adult reti nal ganglion cell survival and neurospheres induces recovery in a chronic model of multi ple axonal regenerati on. Mol Cell Neurosci 2007;34:88-98. sclerosis. Nature 2003;422:688-694. 57. Aguayo AJ, Vidal-Sanz M, Villegas-Pérez MP, Bray GM. Growth and 36. Yu J, Vodyanik MA, Smuga-Ott o K, et al. Induced pluripotent stem connecti vity of axotomized reti nal neurons in adult rats with opti c cell lines derived from human somati c cells. Science 2007;318:1917- nerves substi tuted by PNS graft s linking the eye and the midbrain. 1920. Annals of the New York Academy of Sciences 1987;495:1-9. 37. Mellough CB, Cui Q, Harvey AR. Treatment of adult neural progenitor 58. Aguayo AJ, Bray GM, Carter DA, et al. Regrowth and connecti vity cells prior to transplantati on aff ects graft survival and integrati on of injured central nervous system axons in adult rodents. Acta in a neonatal and adult rat model of selecti ve reti nal ganglion cell Neurobiol Exp (Wars) 1990;50:381-389. depleti on. Restor Neurol Neurosci 2007;25:177-190. 59. Dezawa M, Kawana K, Negishi H, Adachi-Usami E. Glial cells in 38. Pluchino S, Gritti A, Blezer E, et al. Human neural stem cells degenerati ng and regenerati ng opti c nerve of the adult rat. Brain Res ameliorate autoimmune encephalomyeliti s in non-human primates. Bull 1999;48:573-579. Ann Neurol 2009;66:343-354. 60. Sievers J, Bamberger C, Debus OM, Lucius R. Regenerati on in the 39. Pluchino S, Zanotti L, Rossi B, et al. Neurosphere-derived multi potent opti c nerve of adult rats: infl uences of cultured astrocytes and precursors promote neuroprotecti on by an immunomodulatory opti c nerve graft s of diff erent ontogeneti c stages. J Neurocytol mechanism. Nature 2005;436:266-271. 1995;24:783-793. 40. Crisan M, Yap S, Casteilla L, et al. A perivascular origin for 61. Girard C, Bemelmans A-P, Dufour N, et al. Graft s of brain-derived mesenchymal stem cells in multi ple human organs. Cell Stem Cell neurotrophic factor and neurotrophin 3-transduced primate 2008;3:301-313. Schwann cells lead to functi onal recovery of the demyelinated mouse spinal cord. J Neurosci 2005;25:7924-7933. 41. Woodbury D, Schwarz EJ, Prockop DJ, Black IB. Adult rat and human bone marrow stromal cells diff erenti ate into neurons. J Neurosci Res 62. Hausmann B, Sievers J, Hermanns J, Berry M. Regenerati on of axons 2000;61:364-370. from the adult rat opti c nerve: infl uence of fetal brain graft s, laminin, and arti fi cial basement membrane. J Comp Neurol 1989;281:447- 42. Sanchez-Ramos JR. Neural cells derived from adult bone marrow and 466. umbilical cord blood. J Neurosci Res 2002;69:880-893. 63. Ruitenberg MJ, Vukovic J, Sarich J, et al. Olfactory ensheathing cells: 43. Bertani N, Malatesta P, Volpi G, et al. Neurogenic potenti al of human characteristi cs, geneti c engineering, and therapeuti c potenti al. J mesenchymal stem cells revisited: analysis by immunostaining, ti me- Neurotrauma 2006;23:468-478. lapse video and microarray. J Cell Sci 2005;118:3925-3936. 64. Negishi H, Dezawa M, Oshitari T, Adachi-Usami E. Op ti c nerve 44. Wu S, Suzuki Y, Ejiri Y, et al. Bone marrow stromal cells enhance regenerati on within arti fi cial Schwann cell graft in the adult rat. Brain diff erenti ati on of cocultured neurosphere cells and promote Res Bull 2001;55:409-419. regenerati on of injured spinal cord. J Neurosci Res 2003;72:343-351. 65. Chen DF, Jhaveri S, Schneider GE. Intrinsic changes in developing 45. Uccelli A, Mancardi G. Stem cell transplantati on in multi ple sclerosis. reti nal neurons result in regenerati ve failure of their axons. Proc Natl Curr Opin Neurol 2010;23:218-225. Acad Sci USA 1995;92:7287-7291. 46. Yiu G, He Z. Glial inhibiti on of CNS axon regenerati on. Nat Rev 66. Goldberg JL, Klassen MP, Hua Y, Barres BA. Amacrine-signaled loss Neurosci 2006;7:617-627. of intrinsic axon growth ability by reti nal ganglion cells. Science 47. Wong EV, David S, Jacob MH, Jay DG. Inacti vati on of myelin- 2002;296:1860-1864. associated glycoprotein enhances opti c nerve regenerati on. J 67. Moore DL, Blackmore MG, Hu Y, et al. KLF family members regulate Neurosci 2003;23:3112-3117. intrinsic axon regenerati on ability. Science 2009;326:298-301.

2011 Annual Meeting Syllabus | 105 68. Cai D, Deng K, Mellado W, et al. Arginase I and polyamines act downstream from cyclic AMP in overcoming inhibiti on of axonal growth MAG and myelin in vitro. Neuron 2002;35:711-719. 69. Hammarlund M, Nix P, Hauth L, et al. Axon regenerati on requires a conserved MAP kinase pathway. Science 2009;323:802-806. 70. Park KK, Liu K, Hu Y, et al. Promoti ng axon regenerati on in the adult CNS by modulati on of the PTEN/mTOR pathway. Science 2008;322:963-966. 71. Deiner MS, Kennedy TE, Fazeli A, et al. Netrin-1 and DCC mediate axon guidance locally at the opti c disc: loss of functi on leads to opti c nerve hypoplasia. Neuron 1997;19:575-589. 72. Raper JA. Semaphorins and their receptors in vertebrates and invertebrates. Curr Opin Neurobiol 2000;10:88-94. 73. Erskine L, Williams SE, Brose K, et al. Reti nal ganglion cell axon guidance in the mouse opti c chiasm: expression and functi on of robos and slits. J Neurosci 2000;20:4975-4982. 74. Williams SE, Mann F, Erskine L, et al. Ephrin-B2 and EphB1 mediate reti nal axon divergence at the opti c chiasm. Neuron 2003;39:919- 935. 75. Knöll B, Isenmann S, Kilic E, et al. Graded expression patt erns of ephrin-As in the superior colliculus aft er lesion of the adult mouse opti c nerve. Mech Dev 2001;106:119-127.

106 | North American Neuro-Ophthalmology Society LASER-ASSISTED TRANSPLANTATION OF STEM CELLS INTO THE ADULT EYE

Yaping Joyce Liao1, Yi-Wen Chen1, James Weimann1, Madison Stanford1, Joyce Ho1, Lawrence Recht1 Stanford Stanford, CA, United States1

KEY WORDS RESULTS 1. Stem Cell Reti nal lasering led to a dramati c increase in the number of GFP-positi ve ES-NPCs at all ti me points. Lasering of 2. AION twelve 100-micron spots at 300 mW was suffi cient and 3. Laser Photocoagulati on most effi cacious compared to 3 spots or lower power. Two months following transplantati on, many GFP-positi ve ES- 4. Reti nal Ganglion Cells NPCs survived and integrated predominantly into the inner reti na, expressing neuronal marker beta-tubulin or glial marker glial fi brillary acidic protein. Many ES-NPCs formed INTRODUCTION dense clusters which extended beta-tubulin-positi ve Regenerati ve approach to treat anterior ischemic processes that coalesce to form bundles that projected into opti c neuropathy is of sound rati onale and signifi cant the posterior pole. clinical demand because the central nervous system has limited repair and regenerati ve capacity and there are no eff ecti ve neuron- or nerve-protecti ve methods CONCLUSION to treat this devastati ng conditi on. However, stem cell Our data provided strong evidence that reti nal laser transplantati on into the adult eye to reconsti tute neurons photocoagulati on facilitated the intravitreal transplantati on and neural circuitry is diffi cult and oft en of low yield. In of a large number of ES-NPCs into the adult mammalian our study, we tested the potenti al benefi t of reti nal laser inner reti na. This form of controlled injury provided a photocoagulati on, a clinically proven form of controlled permissive environment for stem cell diff erenti ati on into injury used to treat pati ents with diabeti c reti nopathy, on the neuronal pathway and formati on of nerve-like structure intravitreal stem cell transplantati on into the adult eye. within the inner reti na.

METHODS FINANCIAL DISCLOSURE We performed unilateral reti nal laser photocoagulati on Burroughs Wellcome Foundati on in 130 adult wildtype 129 mice followed by bilateral intravitreal injecti on of enhanced green fl uorescent protein (GFP)-expressing neural progenitor cells derived from murine embryonic stem cells (ES-NPCs). Diff erent laser parameters were tested to assess opti mal conditi ons for transplantati on. We analyzed the reti na of lasered vs. unlasered eyes following transplantati on at 2-weeks, 1-month, and 2-months with immunohistochemistry and morphometric analysis.

2011 Annual Meeting Syllabus | 107 108 | North American Neuro-Ophthalmology Society THE FIELD OF VISUAL PROSTHETIC DEVICES: CONCEPTUAL FOUNDATION, STATUS AND CONTROVERSY Joseph Rizzo, III, MD Massachusetts Eye and Ear Infirmary Boston, MA

LEARNING OBJECTIVES are encouraging but also surprising, given the seemingly reasonable assumpti ons that the level of technology is not 1. The att endee will be able to understand the conceptual suffi ciently mature and that reti nal pathologies (following foundati on of the fi eld of visual prostheti c devices. loss of the photoreceptors) would confound att empts to 2. The att endee will develop an appreciati on for the communicate eff ecti vely with the reti na using arti fi cial level of vision that visual prostheti c devices currently sti muli. This conundrum will be discussed, and a blueprint provide to pati ents. of informati on that might be provided to pati ents who inquire about the fi eld of visual prostheti c devices will be 3. The att endee will develop an appreciati on for the type presented. of informati on that might be relevant for pati ents who inquire about whether they should consider having a visual prostheti c implant. Blindness is a major health problem and a common form of disability. In industrialized countries, the most common CME QUESTIONS forms of blindness are the result of neural diseases, either of the reti na or the opti c nerve. Most reti nal forms of 1. What types of blinding conditi ons are amenable to blindness are caused by loss of photoreceptors, and these treatment with a visual prostheti c device? outer reti nal diseases aff ect at least 800,000 United States 2. Which type of visual prostheti c device (eg. reti nal, citi zens each year, roughly 10% of whom are legally blind. opti c nerve, corti cal) is the most favourable for use Most blindness caused by opti c nerve disease is the result in pati ents who are severely blind from reti niti s of glaucoma, which aff ects approximately 1% of the general pigmentosa? populati on. There is no treatment to restore lost vision for any form of neural blindness. 3. What are the factors that would seem to limit the visual potenti al of reti nal prostheti c devices? CONCEPTUAL FOUNDATION FOR VISUAL PROSTHETIC DEVICES A prosthesis is an arti fi cial device that is designed to replace KEYWORDS the functi on of a damaged or lost part of the body. In the 1. Blindness context of a visual prostheti c device, a prosthesis could be uti lized to sti mulate the visual pathway distal to the 2. Visual Prosthesis locati on of the neural damage that caused the blindness. Visual prosthe ti cs are being pursued to be implanted on 3. Reti nal Prosthesis the sub-reti nal or epi-reti nal surface (for outer reti nal 4. Visual Rehabilitati on diseases); around the opti c nerve (for outer or inner reti nal disease); at the lateral geniculate body (for opti c nerve 5. Reti nal Reorganizati on or more proximal disease); and at the visual cortex (for blinding diseases proximal to the primary visual cortex). At present, there is no evidence to support the advantages INTRODUCTION of one type of approach over the other. The concept for a The fi eld of visual prostheti c devices has produced a visual prostheti c device is exactly homologous to the use of remarkable array of sophisti cated technologies over the last cochlear prostheses, which have been remarkably successful 20 years that off er hope for improving the quality-of-life for in restoring hearing to pati ents. Over 150,000 cochlear pati ents who are severely blind from reti niti s pigmentosa. prostheti c devices have been implanted worldwide. Five companies have already performed chronic implants in humans, and all fi ve report fairly positi ve results. In the The hope that a visual prosthesis can create suffi cient vision best case, a legally-blind pati ent (implanted and tested in to improve the quality-of-life for severely blind pati ents is Germany) was able to “read” only one week aft er receiving based upon the generally predictable topographic order a sub-reti nal implant. This and other very positi ve results within and across each neuronal level within the aff erent

2011 Annual Meeting Syllabus | 109 visual pathway. Given this anatomical orderliness, it seems The ease of surgically approaching the eye, and the reasonable to assume that direct electrical sti mulati on of planar and distributed architecture of the neurons in visual neurons might generate phosphenes at locati ons the reti na are some of the reasons that have driven the within the visual fi eld similar to those that would have development of reti nal prosthesti c devices. All but four been obtained by photi c sti mulati on. Similarly, electrical visual prostheti c research programs have chosen to build sti mulati on of the opti c nerve could produce an orderly reti nal (vs. other types of visual) prostheti c devices. Intra- arrangement of phosphenes because of the generally ocular visual prostheti cs are being designed for use in faithful topographical organizati on of opti c nerve axons to the epi-reti nal, sub-reti nal and supra-choroidal spaces, points within the visual fi eld. though only one group is investi gati ng the latt er. Slightly more groups are pursuing the epi- versus the sub-reti nal VARIOUS APPROACHES BEING EXPLORED FOR VISUAL approach. Implantati on within the eye is complicated by PROSTHETIC DEVICES the fact that the reti na is excepti onally delicate (much The fi rst type of visual prostheti c devices to be developed more so that cerebral ti ssue), the eye is prone to persistent were corti cal implants, the fi rst eff ort for which began infl ammati on, and the three dimensional curvature of the in the 1970s.1,2 The only acti ve programs at this ti me reti na presents a challenging substrate for implantati on are based at the Illinois Insti tute of Technology and in of the two dimensional devices that are being developed. Spain (under the directi on of Philip Troyk and Eduardo Failure to implant the sti mulati ng electrode array close to Fernandez, respecti vely). The fi rst real opportunity to (within 10 microns) and conformal to the reti na over a wide create devices that could be made small enough to fi t into enough area (≥ 3mm) could raise the sti mulati on thresholds the eye occurred slightly more than 20 years ago with to an unacceptable level and possibly compromise the availability of customizable microfabricati on methods. quality of induced percepts or the width of the visual fi eld Two reti nal prostheti c groups emerged at roughly the that is generated. Some of these considerati ons are also same ti me, our group at the Harvard Medical School relevant for implants that are positi oned elsewhere along and the Massachusett s Insti tute of Technology, and the the aff erent visual pathway. other at Duke and North Carolina State Universiti es (the latt er eventually became the nucleus of the Second Sight COLLECTIVE ACHIEVEMENTS OF HUMAN PSYCHOPHYSICAL Company). The visual prostheti c fi eld then enjoyed TESTING WITH VISUAL PROSTHETIC DEVICES enormous growth and now includes 22 visual prostheti c The following is a brief summary of the results of human research groups in seven countries. testi ng for each type of visual prostheti c device. At present, reti nal prostheti c devices are the only type of device that is Comparison of diff erent types of visual prostheti c devices being implanted in humans. Each of the locati ons being studied as a site for a visual prosthesis has certain advantages and disadvantages. Visual Corti cal Prostheses For instance, with respect to a visual corti cal prosthesis, In 1967, Brindley implanted an electronically-primiti ve the need for extensive neurosurgery to implant a device device into the visual cortex of a pa ti ent who was would seem to be an impediment, but modern techniques completely blind from glaucoma.1,3 These experiments used for “functi onal electrical sti mulati on” of sites in the and those that followed by Dobelle showed that multi ple deep brain (as is done for Parkinson disease, for instance) phosphenes could be perceived simultaneously following make it possible to implant devices into the brain through sti mulati on of multi ple electrodes, and that there was relati vely small incisions in awake pati ents. Implantati on at a perceptual alignment of the phosphenes that roughly the level of the opti c nerve is best approached by placing a correlated with the spati al organizati on of the visual cuff -style electrode array around the intra-orbital segment cortex.2,3 These experiments were conducted by delivering of the opti c nerve, but approaching the opti c nerve within sti mulati on to the surface of the cortex, which required the orbit is complicated by the fact that at this locati on the using relati vely high charge levels because of the substanti al opti c nerve here is invested by all three meningeal sheaths; distance between the electrode and neurons. Surface thus, the electrodes would lie outside of the dura mater sti mulati on produced only coarse two-point discriminati on which would increase sti mulati on thresholds and make it of phosphenes. The limitati on of these fi ndings moti vated more diffi cult to create highly localized areas of sti mulati on. the development of electrodes that could be implanted into The dura mater and arachnoid are not present along the the visual cortex. Subsequent engineering work, much of intracranial segment of the opti c nerve, but implanti ng which has been performed at the University of Utah and electrodes intracranially would require much more the Illinois Insti tute of Technology, has greatly advanced this extensive surgery. The pursuit of a lateral geniculate body fi eld, but presently no human trials are being conducted (LGB) prosthesis has the advantage that the LGB neurons with these devices. are spati ally and physiologically segregated, but it would be very challenging to implant a large number (i.e. hundreds) of electrodes this deep into the brain.

110 | North American Neuro-Ophthalmology Society Reti nal Prostheses These outcomes would seem to exceed the level of success Five companies (two from the United States; three from that might have been predicted with this populati on of Germany) have performed chronic implants of reti nal end-stage pati ents and with the level of technology that has prostheti c devices in humans. The fi rst company to been available. The following factors each present would perform implants (Optobionics, Inc.) declared bankruptcy seem to argue against some of these successful outcomes: aft er it was understood that the device was not creati ng any visual benefi t based upon electrical sti mulati on from the 1. The presumpti on that a larger number of electrodes, sub-reti nal device. Rather, a “trophic eff ect” was off ered each of which can be controlled independently, would as the basis for the visual improvement that was reported be needed to create spati ally detailed vision; (the by their pati ents. The collecti ve results of the other largest number of individually-controlled channels four groups has provided important “proof-of-concept” currently available is 64); evidence in support of the potenti al value of a reti nal prosthesis, including demonstrati on: 1) of acceptable 2. Pati ents with end-stage reti niti s pigmentosa would not sti mulati on thresholds4,5; 2) that severely blind pati ents can have suffi cient survival of reti nal neurons to enable reliably see phosphenes; 3) that modulati on of electrical spati ally-detailed vision; sti mulati on alters percepts5; 4) that object orientati on and 3. The compelling evidence from Robert Marc that movement can be detected; and 5) that vision of 20/1000 is has shown widespread and profound “reti nal possible, although this usually required head movement to reorganizati on” of the inner reti na following loss of scan lett ers and considerable ti me, someti mes many tens of photoreceptors10-13; seconds, to assimilate the optotypes5,6. The most notable claim, which comes from Tubingen, is of a pati ent who 4. The mismatch between the very positi ve and relati vely could read one week aft er implantati on, which would seem early outcomes with reti nal prostheti c devices and to defy both intuiti on and substanti al scienti fi c evidence. seemingly reasonable factors that would argue against Collecti vely, these results are encouraging but fall short such successes creates a conundrum. How can such of the goal of providing a suffi ciently bett er level of vision successes have occurred so quickly with devices that that is easily perceived to justi fy the risk of surgery and the seem to have inadequate technology for the task? cost of the device compared to whatever strategy a pati ent How can such successes have occurred if the neural might already be using, like a white cane or computer- substrate is so depleted and disorganized? assisti ve technologies. A framework for considering these puzzling factors will be Opti c Nerve Prosthesis presented. Veraart and Delbeke implanted a cuff with four electrodes around the opti c nerve of two pati ents with reti niti s WHAT PERSPECTIVE TO SHARE WITH INTERESTED pigmentosa. Using a camera-based system to capture PATIENTS? visual images, their pati ents were able to judge the shape, The discussion between physician and pati ent on the topic size, basic structure, brightness, locati on and brightness of of visual prostheti c devices is challenging because of the objects. With regard to patt ern recogniti on experiments complexity of the fi eld and because of the vulnerability of basic forms, these researchers reported an 87% success of many severely blind pati ents who long for a cure. rate, but with a relati vely long ti me delay of 53 seconds At the least, it would be enti rely proper to relate that for pati ents to discern and report on the percepti on of visual prostheses: a) are a legiti mate approach for vision objects. With training, the subjects learned to scan more rehabilitati on in selected instances of acquired, neural eff ecti vely and to more quickly discriminate objects on a blindness; b) the fi eld has produced a remarkable array table (e.g. cup, eati ng utensil). The pati ents could use this of very sophisti cated devices; c) that the implant devices informati on to reach out and grasp the objects. Stronger from each group seem to have been well-tolerated by the electrical pulses made the phosphenes appear brighter and eye; d) that the majority of severely blind pati ents reported more central. One pati ent reported being able to perceive at least crude percepts of vision (even without any real a lett er “C”.7-9 training); and e) that in the best cases, unexpectedly good vision, including the ability to read, has been reported. CONUNDRUM Personally, I present these successes together with the Reti nal prostheti c devices reportedly have enabled severely perspecti ve that there is considerable scienti fi c informati on blind pati ents to navigate bett er, recognize lett ers, and even that would seem to suggest that such good vision would to read. Despite the quite diff erent technologies used, each not have been possible, not without further advances both of the four groups that have performed more advanced in technology and a bett er understanding of to deliver psychophysical testi ng (i.e. beyond simple determinati on of electrical sti mulati on to the nerve ti ssue. I explain to my sti mulati on thresholds) has reported success. Every one of pati ents that it is diffi cult to resolve these confl icti ng types these successful outcomes has occurred in a pati ent with of evidence, but that the positi ve outcomes have been end-stage reti niti s pigmentosa. reported by each group that is studying the use of these devices, which provides some reassurance. I also indicate

2011 Annual Meeting Syllabus | 111 that the fi eld is sti ll relati vely early in development, and REFERENCES that it is likely that a much more clear understanding of 1. Brindley GS. Sensati ons produced by electrical sti mulati on of what can be achieved will be realized in the next 5-10 the occipital poles of the cerebral hemispheres, and their use in years. Even higher quality vision may be realized as the constructi ng visual prostheses. Ann R Coll Surg Engl 1970;47:106-8. devices are upgraded. I also discuss other potenti al types 2. Dobelle WH, Mladejovsky MG. Phosphenes produced by electrical of research that might restore vision to the blind, including sti mulati on of human occipital cortex, and their applicati on to the stem cell therapy. I also emphasize the importance of development of a prosthesis for the blind. J Physiol 1974;243:553-76. vision rehabilitati on to maximize a pati ent’s potenti al for 3. Brindley GS. Eff ects of electrical sti mulati on of the visual cortex. Hum vision for whatever level of vision they might have. Neurobiol 1982;1:281-3. 4. Mahadevappa M, Weiland JD, Yanai D, Fine I, Greenberg RJ, Humayun MS. Perceptual thresholds and electrode impedance in three reti nal prosthesis subjects. IEEE Trans Neural Syst Rehabil Eng CME ANSWERS 2005;13:201-6. 1. A visual prosthesis could potenti ally off er a strategy to 5. Yanai D, Weiland J, Mahadevappa M, Greenberg R, Fine I, Humayun restore vision to pati ents with acquired forms of neural M. Visual performance using a reti nal prosthesis in three subjects blindness that aff ected the aff erent visual pathway with reti niti s pigmentosa. Am J Ophthalmol 2007;5:820-827. anywhere from the photoreceptors to the opti c 6. Memon MA RJ. Development of Visual Prostheses and Other radiati ons. Theoreti cally, a visual prosthesis could be Assisti ve Devices. In: Levin L, ed. Ocular Diseases: Elsevier; 2009. used at higher visual corti cal levels, but presently the 7. Veraart C, Wanet-Defalque MC, Gerard B, et al. Patt ern recogniti on fi eld is exploring strategies that depend upon having a with the opti c nerve visual prosthesis. Arti f Organs 2003;27:996- 1004. (nominally) intact primary visual cortex. 8. Veraart C, Duret F, Brelen M, et al. Vision rehabilitati on in the case of 2. No parti cular embodiment of visual prosthesis has blindness. Expert Rev Med Devices 2004;1:139-53. shown consistently more promise or problems than 9. Duret F, Brelen ME, Lambert V, et al. Object localizati on, other devices. At present, only reti nal prostheti c discriminati on, and grasping with the opti c nerve visual prosthesis. devices are being implanted in humans, and both the Restor Neurol Neurosci 2006;24:31-40. epi- and sub-reti nal implants have shown signifi cantly 10. Marc RE, Jones BW. Reti nal remodeling in inherited photoreceptor degenerati ons. Mol Neurobiol 2003;28:139-47. positi ve results. The sub-reti nal device from Reti na Implant AG (based in Tubingen) has been reported 11. Jones BW, Marc RE. Reti nal remodeling during reti nal degenerati on. Exp Eye Res 2005;81:123-37. to allow a blind pati ent to read without the need for head scanning and with a shorter ti me to acquire the 12. Marc RE, Jones BW, Watt CB, et al. Neural remodeling in reti nal degenerati on. Prog Reti n Eye Res 2003;22:607-55. image compared to the most positi ve reports of an epi- reti nal device, though there is no ability to scienti fi cally 13. Marc RE, Jones BW, Anderson JR, et al. Neural reprogramming in reti nal degenerati on. Invest Ophthalmol Vis Sci 2007;48:3364-71. compare these diff erent approaches. 3. The potenti al for reti nal prostheti c devices to restore vision would seem to be limited by at least two factors: 1) pathologies that develop in reti nas following loss of photoreceptors (i.e. “reti nal reorganizati on”); and 2) by the state of engineering development that has yet to produce devices with a large enough number (hundreds, perhaps) of individually-controllable sti mulati on sites using electrodes that can be positi oned close to (within 10 microns, or so) neurons without inducing mechanical or electrical damage to the neurons.

112 | North American Neuro-Ophthalmology Society UTAH ARTIFICIAL VISION PROJECT—OCCIPITAL CORTICAL IMPLANTS FOR BLINDNESS Judith Warner, MD John Moran Eye Center, University of Utah Salt Lake City, Utah

LEARNING OBJECTIVES announced implantati on of surface corti cal sti mulati ng systems in humans, initi ally performed in the US and later 1. The att endee will be able to describe the current status in Portugal. The sti mulati ng electrodes were hooked up to a of the Utah Arti fi cial Vision Project. digital video camera, computer and associated electronics. 2. The att endee will be able to list anatomic constraints The pati ents received corti cal sti mulati on through the and advantages of corti cal implantati on for vision device. One pati ent, Jens Naumann, was widely reported restorati on. in dramati c news stories and videos (3). The sti mulati ng electrodes were on the surface of the cortex, requiring 3. The att endee will be able to describe the moderately intense sti mulati on to elicit a phosphene – 1-10 psychophysical issues of corti cal implants. milliamps. Unfortunately, the system induced seizures and later lost its effi cacy and had to be removed. With Dr. Dobelle’s death in 2004, litt le informati on is now known CME QUESTIONS about the subjects, their percepts, or the long-term 1. How many electrodes are in the Utah Electrode Array? outcomes. In the same year (2000), a single subject at the NIH had intracorti cal sti mulati ng electrodes implanted for 2. What is the percept induced by electrical sti mulati on of a short ti me (4). He died aft er a subarachnoid hemorrhage the occipital cortex called? from an unrelated intracranial aneurysm. NIH briefl y 3. What is the apparent cause of long term corti cal considered conti nuing the project with outside assistance electrode implant recording and sti mulati on but eventually decided not to proceed. degradati on? The Utah Arti fi cial Vision Project (UAVP) is a multi disciplinary eff ort to explore and implement the use KEYWORDS of direct occipital lobe sti mulati on to provide useful visual informati on to blind humans. The project has involved 1. Arti fi cial Vision parti cipants from many departments of the University 2. Utah Electrode Array of Utah. These include bioengineering, neuro-radiology, neurosurgery, and ophthalmology. Richard Normann PhD 3. Occipital Cortex is the director of the project, and Bradley Greger PhD is the director of primate studies. Some human subjects have also 4. Corti cal Implants been parti cipants in these eff orts, both in the laboratory 5. Phosphene and in the surgical suite.

The strategy of the UAVP is to translate real-ti me video INTRODUCTION into patt erned sti muli and directly sti mulate the visual cortex with penetrati ng electrodes (5,6). This strategy takes The Utah Arti fi cial Vision Project (UAVP) is a advantage of the occipital cortex foveal magnifi cati on and multi disciplinary eff ort to explore and implement the use of its reti notopic organizati on. The device developed has direct occipital lobe sti mulati on in order to provide useful been named the Utah Electrode Array (UEA). The Utah visual informati on to blind humans. There are numerous Electrode Array is built from silicon and is a cluster of advantages to the strategy of sti mulati on of the occipital 100 microelectrodes. The sti mulati ng ti p has a sputt ered lobe, but also numerous technical hurdles. Early results iridium oxide fi lm. This array is designed to be implanted from primates and human support the validity of the into the surface of the occipital cortex, to provide direct concept. sti mulati on to the underlying corti cal neurons in patt erns directed by the real ti me video apparatus. Each electrode In 1968, Brindley (1) showed that electrical sti mulati on of communicates with 2 to 3 neurons. The electrodes are the visual cortex resulted in percepti on of a point of light powered and controlled remotely. Dr. Normann esti mates known as a phosphene. In 2000, Dobelle (2) (who was a 6 UEAs could restore useful vision sense to those with graduate student at the University of Utah in the 1970s) profound blindness. The advantage of the strategy of

2011 Annual Meeting Syllabus | 113 occipital sti mulati on, as opposed to reti nal or opti c nerve The anatomic benefi t of the occipital pole is the sti mulati on, is that it can be used even in pati ents with magnifi cati on of the reti notopic map, from a small, densely severe injury to the anterior visual system. Therefore, packed area of the reti na, to the largest area of the it could be used in pati ents aft er enucleati on or with occipital cortex. Thus, theoreti cally, Dr. Normann believes severe reti nal or opti c nerve damage. Normann’s group that UEA sti mulati on of the neurons of the occipital pole believe that the UEA will be more successful than the could provide high-resoluti on vision with a relati vely low- Dobelle system because of substanti ally lower sti mulati on resoluti on electrode array, our most sought-aft er goal in requirements (1-10 microamps instead of milliamps) due to vision restorati on. penetrati on of the electrodes. One of our former residents, Paul Yang, is working on Other laboratories around the country and around the the issue of corti cal changes that might be induced aft er world are studying the concept of arti fi cial vision. One, prolonged blindness. He is addressing the questi on of also using the occipital cortex, is the laboratory of Dr. Philip identi fying and assessing the occipital cortex in pati ents Troyk at the Illinois Insti tute of Technology. This group with recent versus long-term blindness. includes many members of the short lived NIH human experiment. The device studied at IIT has penetrati ng LONG TERM IMPLANTS AND ELECTRODE SURVIVAL electrodes similar to the UEA, but it is smaller and has In the early 2000’s Dr. Normann collaborated with our fewer electrodes. His most recent directi on of research neurosurgical department (15,16) to perform histological appears to be a wireless electrode array (7). Other labs are studies of corti cal ti ssue subjected briefl y to the working on sti mulati ng the reti na, opti c nerve and lateral implantati on of the UEA. Pati ents scheduled to undergo geniculate nucleus. corti cectomy for intractable epilepsy agreed to have the device implanted into ti ssue that was then removed and PSYCHOPHYSICS examined. The implantati on procedure was adjusted to A great deal of ti me has been devoted to the questi on minimize corti cal deformity and microhemorrhages. The of what input to the visual cortex is necessary to provide devices were able to record single neurons from these useful vision. Psychophysical experiments on graduate human subjects. students in the 1990s used volunteers who att empted to navigate a maze and read using Arti fi cial Vision goggles The devices have been implanted in cats and in a single simulati ng pixelized vision. Most researchers have agreed non-human primate. The electrodes functi on to record as that 600 to 700 pixels of informati on are needed to well as to sti mulate. In the cats, single neuron recordings navigate and even to read. In additi on, the goggles do not have been made from up to 75 of the microelectrodes take into account the phenomenon of corti cally sti mulated of the UEA for as long as 8 months. On histological phosphenes, which move with movement of the eyes (8,9,10). examinati on of ti ssue, there is fi brosis surrounding the electrode sites, which does not correlate with signal ANATOMIC CONSTRAINTS degradati on. There is also some infl ammati on, which The surgical approach to this endeavour must be borne correlates bett er with degradati on of signal (Normann, in mind. As we are all aware, based on years of study by personal communicati on). great anatomists, including our own Horton and Hoyt, the occipital corti ces occupy the mesial surface of the posterior Primates with approximately 1 year of training can indicate lobes of the brain. This region, encompassing the primary whether or not they perceive a photi c sti mulus on a visual cortex, is challenging to access surgically. Therefore, monitor. They gaze at a central fi xati on target. A second most att enti on has been paid to the porti on of V1 on visual sti mulus of a random locati on, brightness, and size the more surgically friendly outer surface. Fortuitously, appears. Aft er a tone is sounded, the monkey lift s one numerous studies have confi rmed the macroscopic patt ern hand if it saw something and lift s the other hand if it saw of corti cal representati on of the reti na, with the macular nothing. When the occipital cortex is sti mulated with an regions located at the occipital ti ps. Those same studies implanted UEA, the monkeys indicate that they perceive have confi rmed tremendous variability in the amount of phosphenes. However, there are concerns that only the total V1 exposure. Dobelle and colleagues (11) conducted a minority of the microelectrodes evoke a response. This study of the locati on of the line of Gennari in 1974, with may be because the monkeys are too discriminati ng in 52 cadaver hemispheres fi xed in formaldehyde, and found what they deem worthy of “calling” a percept. They may a four fold variati on from 359 to 1208 mm2, with 1/3 on not respond to sti muli that evoke a percept that is too the mesial surface, and 3% exposed on the occipital pole. large, an untested colour or shape, etc. It may also be that These authors found the average exposure on the occipital the implant is too distant from the majority of neurons, pole was 68 mm2 with a range 0-260 mm2 (Stensaas et al having migrated away from the corti cal surface (Normann, 1974). More recently, studies using high-resoluti on 3T MRI personal communicati on). in-vivo have shown that we can identi fy the line of Gennari to map V1 (12,13,14) with reasonable accuracy in vivo in just over half of individuals.

114 | North American Neuro-Ophthalmology Society CONCLUSION REFERENCES As we speak, permission is being sought from the FDA to 1. Brindley GS, Lewin WS. The sensati ons produced by electrical implant these devices in human volunteers for 30 days. sti mulati on of the visual cortex. J Physiol. 1968 May;196(2):479-93. There are many questi ons to be answered before and 2. Dobelle WH. Arti fi cial vision for the blind by connecti ng a television aft er human implantati on occurs. As in animals, long-term camera to the visual cortex. ASAIO J. 2000 Jan-Feb;46 (1):3-9. electrode stability and phosphene thresholds will have 3. Kotler S. Vision Quest. A half century of arti fi cial-sight research has to be determined. Criti cally, can complex spati al percepts succeeded, and now this blind man can see. Wired, 2002( Archive be created from phosphenes? How will neighbouring online 10.09). electrodes be perceived when sti mulated individually or 4. Bak M, Girvin JP, Hambrecht FT, et al. Visual sensati ons produced by separately? What adaptati on can occur over ti me, both in intracorti cal microsti mulati on of the human occipital cortex. Med “making sense” of new percepts, and to the moti on of the Biol Eng Comput. 1990 May;28(3):257-9. percepts induced by eye movement? New electrode arrays 5. Normann RA, Maynard EM, Rousche PJ, et al. A neural interface for a that sti mulate diff erent layers of the cortex are also being corti cal vision prosthesis. Vision Res. 1999 Jul;39 (15):2577-87. studies. 6. Maynard EM, Nordhausen CT, Normann RA. The Utah intracorti cal Electrode Array: a recording structure for potenti al brain- computer interfaces. Electroencephalogr Clin Neurophysiol. 1997 Mar;102(3):228-39. ACKNOWLEDGEMENTS 7. Troyk PR, Bradley D, Bak M, et al. Intracorti cal visual prosthesis Much of the material in this syllabus chapter is taken research - approach and progress. Conf Proc IEEE Eng Med Biol Soc. from personal communicati on with Richard Normann 2005;7:7376-9. PhD. I would like to thank him for his ti me. Several of the 8. Cha K, Horch KW, Normann RA. Mobility performance with a references were provided by Paul Yang, MD. Both of them pixelized vision system. Vision Res. 1992 Jul;32 (7):1367-72. have generously supplied me with several Powerpoint 9. Cha K, Horch KW, Normann RA, et al. Reading speed with a pixelized presentati ons from which to fi lch slides and informati on. vision system. J Opt Soc Am A. 1992 May;9 (5):673-7. 10. Cha K, Horch K, Normann RA. Simulati on of a phosphene-based visual fi eld: visual acuity in a pixelized vision system. Ann Biomed Eng. 1992;20(4):439-49. CME ANSWERS 11. Stensaas SS, Eddington DK, Dobelle WH. The topography and variability of the primary visual cortex in man. J Neurosurg. 1974 1. 100 Jun;40 (6):747-55. 2. Phosphene 12. Barbier EL, Marrett S, Danek A, et al. Imaging corti cal anatomy by high-resoluti on MR at 3.0T: detecti on of the stripe of Gennari in 3. Reacti ve Infl ammati on visual area 17. Magn Reson Med. 2002 Oct;48 (4):735-8. 13. Walters NB, Egan GF, Kril JJ, et al. In vivo identi fi cati on of human corti cal areas using high-resoluti on MRI: an approach to cerebral structure-functi on correlati on. Proc Natl Acad Sci U S A. 2003 Mar 4;100(5):2981-6. Epub 2003 Feb 24. 14. Bridge H, Clare S, Jenkinson M, et al. Independent anatomical and functi onal measures of the V1/V2 boundary in human visual cortex. J Vis. 2005 Feb 11;5 (2):93-102. 15. Normann RA, Greger B, House P et al. Toward the development of a corti cally based visual neuroprosthesis. J Neural Eng. 2009 Jun;6(3):035001. Epub 2009 May 20. 16. House PA, MacDonald JD, Tresco PA, et al. Acute microelectrode array implantati on into human neocortex: preliminary technique and histological considerati ons. Neurosurg Focus. 2006 May 15;20(5):E4.

2011 Annual Meeting Syllabus | 115