Doc Ophthalmol (2014) 129:13–55 DOI 10.1007/s10633-014-9441-2

ABSTRACTS

The 52nd ISCEV International Symposium Abstract Issue



Springer-Verlag Berlin Heidelberg 2014

Adachi award lecture High standards in retina and art Michael F. Marmor Department of Ophthalmology, Stanford School of Medicine, Palo Alto, CA, USA ISCEV is the world guardian of standards for clinical electrophysiology—but these did not appear out of nowhere, or without some contention. My role in this process, and pride in its outcome, reflects both science and practicality. However, if the retina is worthy of high standards in testing, it is surely worth pondering ‘‘Why?’’ Because what we test, is what we see. And seeing light and dark and color and contrast are the code that makes retina different from a camera, and art different from photography. Artists use retinal physiology without thinking…as do us mortals, as viewers of art. But a few artists were not so lucky, when vision went awry. This talk will take ISCEV history to vision and great art—hopefully, with respect for all.

William Dawson memorial lecture Gene therapy consequences in patients with RPE65LCA and in animal models Artur V. Cideciyan Scheie Eye Institute, University of Pennsylvania, Philadelphia PA, USA Gene therapy holds great promise in treating inherited retinal degenerations which encompass varied diseases caused by more than 200 different genes. The first such genetic disease treated was Leber congenital amaurosis (LCA) due to RPE65 mutations. Three trials started in 2007, and now there are seven independent gene therapy trials ongoing for RPE65-LCA.

What is the reason for all this attention to a rare condition? What are the long-term effects of gene therapy on retinal function and structure in patients, and in animal models? What are the next steps? RPE65 is the retinoid isomerase within the RPE-based visual cycle which is required for the continuous generation of the 11-cis-retinal chromophore used to capture photons in rod and cone outer segments. Mutations in RPE65 result in a complex interaction of two pathological mechanisms—dysfunction and degeneration of RPE and photoreceptors. There is severe blindness that starts either congenitally or in early childhood, and progresses over years to decades. We examined 15 patients serially over an average of 5 years to better understand the natural history of the dysfunction and degeneration components of the disease. We found that over this period there is detectable thinning of the outer nuclear layer (ONL), and the kinetics of progression are consistent with an exponential decay function. Natural history of visual sensitivity measured at the same loci over the same period showed unexpectedly shallower rate of progression compared to the underlying retinal degeneration. Murine and canine models of RPE65LCA showed the dysfunction component of the disease at their earliest ages to be similar to that in patients. The initiation of the retinal degeneration component, however, was delayed to beyond 5 dog years (equivalent to [30 human years) and beyond 4 mouse months (equivalent to [12 human years); this was unlike the patients who showed substantial retinal degeneration already by 3 human years. We performed gene augmentation therapy in 12 patients with RPE65-LCA, and we evaluated the consequences for up to 3 years after therapy. Patients showed large improvements in both rod and cone photoreceptor mediated function at treated retinal loci within weeks to months after the treatment, and these early gains persisted through the most recent time point evaluated. Unexpectedly, the retinal degeneration measured at the same retinal locations showing visual improvements continued to advance along a trajectory consistent with the natural history (of -0.04 log per year or -9.6 % per year). Gene therapy

123

14 neither accelerated nor arrested the rate of the retinal degeneration. RPE65-mutant dogs treated with the same gene therapy vector during the dysfunction-only stage of their disease showed improved function and protection of treated photoreceptors from degeneration for up to 11 years. Dogs treated at a later disease stage with a human-like dysfunction and degeneration however showed the unabated continuation of photoreceptor loss even though there was improvement in function. The results to date suggest that retinal disease resulting from RPE65 mutations in humans, dogs and mice are complex; comparisons should be done carefully but to an initial approximation dogs [6 years of age and mice [12 months of age are most similar to young human children with RPE65LCA. The consequences of gene augmentation therapy in different species are also complex, and they suggest that the presence of visual function improvement does not imply protection from retinal degeneration. There is a need for combinatorial strategy in RPE65-LCA in order to slow retinal degeneration in the long term in addition to the improvement of function successfully achieved by gene augmentation.

PA1: acquired retinal disease Monday, July 21 9:00–10:30 am

(9:00) Flash ERG in ocular ischemic syndrome S. Jalali, J. Myneni, C. Sahu, D. Reddy L. V. Prasad Eye Institute, Hyderabad, India Purpose To evaluate the flash ERG changes in patients with ocular ischemic syndrome (OIS). Methods Flash ERG findings of patients diagnosed with Ocular Ischemic Syndrome from January 2001 to January 2011 were retrospectively analyzed. The scotopic maximal combined response (MCR) b-wave amplitude and implicit times, b/a ratio, presence or absence of oscillatory potentials and the photopic 30 Hz flicker b- wave implicit time were analyzed as markers of ischemia. Results Twenty-eight patients (56 eyes) diagnosed clinically with OIS had undergone ERG testing in the study period. Mean age was 57.2 ± 13.8 years. There were 18 males and 10 females. The mean scotopic MCR b-wave amplitude (normal = 260 lV) was 190.43 ± 126.77 lV, the median was 207.5 lV and the range was 0–536 lV. The mean MCR b-wave implicit time (normal = 49.4 ms) was 56.7 ± 20.6 ms, the median was 60.65 ms and the range was 0–101 ms. The mean b/a ratio was 2.53 ± 1.77. The mean photopic 30 Hz flicker b-wave implicit time (normal = 34.1 ms) was 36.3 ± 14.6 ms, the median was 40.9 ms and the range was 0–50.9 ms. Oscillatory potentials were absent in 39 out of the 56 eyes. Prolonged b-wave implicit times and absent oscillatory potentials were noted in the majority of cases. In those with advanced disease, the amplitude of the b-wave was low. In those cases with sufficient choroidal vascular compromise and resultant outer retinal ischemia, the a-wave amplitudes were also diminished. This seems to be consistent with the compromise of both retinal and choroidal perfusion in patients with OIS.

123

Doc Ophthalmol (2014) 129:13–55 Conclusions Flash ERGs can be done to complement the other diagnostic modalities and detect widespread retinal ischemia in various stages of OIS. Prolonged b-wave maximal implicit times, prolonged 30-Hz flicker b-wave implicit times and altered b/a wave ratios were significantly different in OIS eyes as compared to controls. The advantages of using the ERG to detect ocular ischemia include non-invasiveness, insignificant effects of cataract, and detection of global ischemia and inner retinal ischemia that are not detected on routine fundus fluorescein angiography. (9:15) Assessment of visual functions in Behc¸et’s disease patients with electroretinogram B. Lei, X. Xiao, P. Yang Department of Ophthalmology, The First Affiliated Hospital of Chongqing Medical University, Chongqing Key Laboratory of Ophthalmology, Chongqing Eye Institute, Chongqing, China Purpose To investigate whether the electroretinogram (ERG) provides a reliable means to evaluate retinal function in Behc¸et’s disease (BD) patients. Methods Full-field ERGs in a cohort of one hundred and ten BD patients (two hundred and twenty eyes) were analyzed retrospectively. All patients underwent routine ophthalmic examinations. The patients (eyes) were divided into groups on the basis of epidemiological data (gender, age of onset, disease duration), ocular clinical features (visual acuity and funduscopic examination) and systemic signs (genital ulceration, multiform skin lesions and arthritis). Dark- and light-adapted ERGs were conducted following the ISCEV ERG standard. We compared the parameters of the dark- and light-adapted ERG among the groups. Results The dark- and light-adapted ERG a- and b-wave amplitudes were proportional to visual acuity. Both the cone and the rod systems were affected in BD patients, but the rod system was more vulnerable than the cone system. Compared to the photoreceptor cells, the inner retinal functions appeared more affected, as indicated by decreased b/a ratio and OP responses. The duration of the disease had an adverse impact on the ERG amplitudes: the 10–15 year group and the 15–20 year group presented significantly decreased responses as compared to those of the \1 year group. The amplitude of b-wave in patients with retinal atrophy was significantly lower than those without retinal atrophy. The systemic signs and the onset age of disease had no impact on ERGs. Conclusions ERG examination provides a reliable assessment for retinal function in Behc¸et’s disease. In agreement with the pathophysiology of the disease, our data suggest that the rod system and the inner retina are primarily affected. (9:30) Diabetic Retinopathy Screening Using RETeval, a Non-Mydriatic 30 Hz Flicker ERG Device Utilizing Skin Electrodes Q. Davis1, A. Maa2, E. Pillow3, W. Feuer4, T. Brown3, R. Caywood3, S. Fransen5 LKC Technologies, Inc., Gaithersburg, MD; 2Atlanta VA Medical Center, Atlanta, GA,; 3Oklahoma VA Medical

1

Doc Ophthalmol (2014) 129:13–55 Center, Oklahoma City, OK; 4Bascom Palmer Eye Institute, University of Miami Health System, Miami, FL; Dean McGee Eye Institute, Oklahoma City, OK, USA Purpose This study measures the performance of the RETeval device’s 30 Hz flicker ERG test to detect vision-threatening diabetic retinopathy (DR). Methods Diabetic patients were selectively recruited in this institutional review board (IRB) approved study at two Veteran Affairs (VA) medical centers in the United States. The planned oversampling attempted to obtain 80 patients for each of five levels of retinopathy/macular edema. For each subject, the RETeval test was performed followed by an Amsler grid test and then, on about 25 % of the subjects, the RETeval test was repeated to assess test precision. Patients were then dilated with tropicamide and phenylephrine eye drops and Early Treatment Diabetic Retinopathy Study (ETDRS) compliant 7 field stereo fundus photographs were taken. These photographs were doubleread in a masked fashion at a dedicated reading center (Inoveon). Photographic results which differed by more than 1 ETDRS step were sent to adjudication where two readers and a retinal specialist conferred to determine the final ETDRS designation. Results 408 patients completed the study. Demographics were 354 male, 54 female, 206 Caucasian, 172 African Americans, 16 Native Americans, 9 Hispanic, 1 Asian, and 4 ‘‘other’’. The majority of patients, 390, were Type II diabetics and 18 were Type I diabetics. The RETeval device had a technical failure rate (no results generated) of 0.7 % (3/408 patients) whereas ETDRS fundus photography (ungradable images) had a significantly higher (p \ 0.001, exact McNemar test) technical failure rate of 15 % (60/408 patients). Two RETeval failures occurred when the device could not detect a pupil (both eyes in one subject and one eye in the other) and the third failure occurred when the device had a poor electrical measurement. The subject where the device could not detect either pupil had sub-1 mm pupils. This individual only dilated to 3.5 mm, which limited the photography, but the RETeval device was able to generate results on that particular subject after dilation. Conclusions The RETeval device shows promise as a new DR screening tool. It has several advantages over current clinical methods. Testing time with the RETeval device is only a few minutes per patient, with no dilation required. Testing was performed by a technician with minimal training. Compared to fundus photography, the RETeval device demonstrated a very low technical failure rate. Funding NIH 9R44EY021121. The content is solely the responsibility of the authors and does not necessarily represent the official views of the sponsor. Commercial relationship: Q. Davis is an employee of LKC Technologies, Inc, which sells the RETeval device. The other authors have no commercial interest in the work.

(9:45) Abnormal full-field electroretinogram after vitrectomy in eyes with idiopathic epiretinal membranes A. Tanikawa, I. Ueda, R. Sakurai, T. Mizutani, T. Mizuguchi, A. Nakamura, Y. Shimada, M. Horiguchi Department of Ophthalmology, Fujita Health University School of Medicine, Toyoake Aichi, Japan

15 Purpose To compare pre- and post-operative full-field ERGs in eyes with idiopathic epiretinal membranes (ERM). Methods The ERGs were elicited from ten eyes of ERM with skin electrodes, by using pulse reference power line noise reduction system (PuREC, Mayo, Aichi Japan). The amplitude of each component in the ERG waveform was compared before, 1 day and 1 week after surgery. Transconjunctival sutureless 25-gauge vitrectomy was performed with phacoemulsification and IOL implantation. The epiretinal membrane was removed, followed by internal limiting membrane peeling without any staining. Results There were no serious complications in any of the cases. Relative amplitudes (mean ± SE), to the baseline, 1 day and 1 week after surgery of each component were scotopic ERG: 85.4 ± 9.8 and 117.6 ± 9.5 %, maximal response a-wave: 89.2 ± 9.6 and 108.1 ± 6.7 %, maximal response b-wave: 80.8 ± 7.2 and 110.9 ± 6.1 %, photopic ERG a-wave: 95.6 ± 13.4 and 123.5 ± 11.5 %, photopic ERG b-wave: 93.4 ± 6.7 and 127.9 ± 8.1 %, photopic negative response (PhNR): 86.8 ± 10.5 and 67.7 ± 7.7 %, flicker ERG: 89.8 ± 7.5 and 98.3 ± 9.2 %, respectively. Conclusions All ERG components showed a reduction in amplitude on the first post-operative day. All except the PhNR recovered 1 week after surgery but the PhNR revealed further reduction. These results suggest that there are some functional impairments in the inner retina after ERM surgery.

(10:00) Systemic indicators for complications of diabetes affect the mfERG T. Wright, R. Dragas, W. Tan, C. Westall Neuroscience and Mental Health, Hospital for Sick Children; Optometry and Vision Science, University of Waterloo; Department of Ophthalmology and Vision Science, University of Toronto, Canada Purpose To describe how the mfERG is affected by systemic markers of complications associated with diabetes. Complications of diabetes include cardiovascular problems and nephropathy as well as diabetic retinopathy (DR). Dyslipidemia is a risk factor for cardiovascular problems, systemic levels of cholesterol are used to diagnose dyslipidemia. Levels of adenine and creatinine in the blood are measured and reported as a ratio (ACR) as a marker for kidney health. These markers are routinely collected by the endocrinology department. The mfERG has been reported as abnormal in patients with diabetes before the onset of DR. Methods A retrospective chart review of adolescent patients with type 1 diabetes and no (or mild) diabetic retinopathy (DR) who had undertaken a standard (103 hexagon) mfERG as part of a previous study. Data collected included resting blood pressure, HbA1c levels, fasting cholesterol levels [low density lipoproteins (LDL), high density lipoproteins (HDL) and triglycerides] and ACR. Measures recorded closest to the mfERG date are reported. A mixed model design is used to assess correlation between mfERG amplitude and the independent variables. Results The amplitude and implicit time of the mfERG P1 components from 52 patients (median age at mfERG test 16 [12–22] years) were identified manually. Blood pressure measurements were available for 100 % of patients (median time from mfERG 1.6 months); cholesterol levels was available for 47 patients (median time from mfERG 11.6 months) and ACR for 20 patients (mean time from mfERG

123

16 14.2 months). 3 patients had LDL above the recommended levels; 5 patients had increased triglyceride levels. There is a positive correlation between mfERG amplitude and levels of HDL (p  0.001) and a negative correlation between mfERG amplitude and LDL (p \ 0.01). Conclusion Changes in blood cholesterol levels correlate with changes to the amplitude of the mfERG. These findings raise the important possibility that dyslipidemia may impact retinal function. (10:15) Predicting progressive visual loss in early AMD by multifocal electroretinogram L. Ambrosio1, G. Nicoletti1, G. de Crecchio1, B. Falsini2, G. Ambrosio1 1

Department of Neuroscience, University of Naples Federico II, Naples; 2Department of Ophthalmology, Catholic University, Rome, Italy

Purpose To investigate the role of multifocal electroretinogram (mfERG) to predict visual acuity decline in early agerelated macular degeneration (AMD). Methods Twenty-six early AMD patients (12 males and 14 females, mean age: 66.9 + 9.8; range 46–82 years) were included in the study. Visual acuity testing with a calibrated standard Snellen and ETDRS charts, OCT (Spectralis OCT, Heidelberg Engineering, Germany), microperimetry (NIDEK, MP1 Microperimeter) and mfERG (Retiscan, Roland Germany) were performed at the study entry (baseline) and after 12 months. mfERG testing was performed according to ISCEV Standard for clinical multifocal electroretinography. The first-order kernel mfERG responses were analyzed. Individual mfERG responses for the 103 hexagons were grouped into six concentric rings centered on the fovea for analysis. The amplitude densities (ADs) of the first positive peak (P1, nV/deg2) for Ring 1–6 were considered. At baseline and after 12 months, foveal retinal thickness (lm) and mean retinal sensitivity (dB) were evaluated for each subject. Results After 12 months, the decline of visual acuity evaluated with ETDRS charts was significantly larger (2.5 letters vs 1 letter, p = 0.026) in patients whose mfERG P1 response AD of Ring 1 at baseline was\100 nV/deg2. The mfERG P1 amplitude of Ring 1 was inversely correlated with the age of the subjects (r = -0.4, p = 0.04). mfERG P1 AD of Ring 1 was positively correlated with mean retinal sensitivity (r = 0.45, p = 0.021). Conclusions The present results indicate that mfERG P1 response AD of Ring 1 provides a good biomarker for detecting early retinal dysfunction and predicting visual acuity decline in early AMD.

Doc Ophthalmol (2014) 129:13–55 2

Centre for Systems Neuroscience, University of Leicester, Leicester, UK

Purpose The classical technique used to enhance the signal-tonoise ratio (SNR) of VEPs involves increasing the number of averaged sweeps, an effective but time-intensive approach. This technique relies on the assumption that responses do not vary from trial to trial. The purpose of this study is to evaluate the value of a method to improve the SNR of averaged VEPs and allow the analysis of single-trial responses using the wavelet transform (WT). Methods To date, 30 checkerboard pattern reversal VEPs were recorded on two normal subjects following ISCEV standard procedures. Averaged responses were de-noised using an automatic method based on the WT (Ahmadi M & Quian Quiroga R, Neuroimage 2013; 66: 627–680). SNR was calculated as a ratio between root-mean-square amplitude values of a signal window (60–160 ms post-stimulus) and a signal-free window (250–350 ms post-stimulus) (Zhang et al. Doc Ophthalmol 2002; 104: 287–302). Results SNR improvement obtained using the WT method was compared with that obtained by doubling the number of sweeps (from n = 100 to n = 200) (classical method). For VEPs recorded at the Oz electrode, the classical method yielded an enhanced SNR, as expected, from a median value of 5.13 for n = 100 to a median value of 6.81 for n = 200 (p \ 0.001, Wilcoxon signed rank test—WSR). In contrast, a much larger SNR enhancement was obtained using the WT method, to a median value of 796.42 for de-noised (n = 100) waveforms (p \ 0.00005, WSR). Obviously, the difference between the classical and WT methods was highly significant (p \ 0.000006, WSR). In addition, single-trial responses could be detected. The analysis of VEPs recorded at other recording sites [O1, O2, POz, and Iz (inion)] yielded comparable results. Conclusions The automated WT de-noising method provided a very efficient approach to increase VEP SNR, achieving a much greater enhancement, with a briefer recording time, than the classical method. The analysis of single-trial responses allowed a measurement of variability within a recording and the calculation of latency-corrected averages. Other ISCEV responses (small amplitude ERGs, mfERGs, PERGs, mfVEPs) would also benefit from this approach.

(11:15) The steady-state VEP as a superposition of single-stimulus responses S. P. Heinrich, M. Groten, M. Bach

PA2: Methods 1 Monday, July 21, 11:00 am–12:30 pm

(11:00) De-noising the ISCEV standard VEP using wavelet transforms R. C. Caruso1, R. Q. Quiroga2, M. Ahmadi2, S. Kastner1 1

Princeton Neuroscience Institute and Department of Psychology, Princeton University, Princeton, NJ, USA;

123

University Eye Center Freiburg, Freiburg, Germany Purpose Steady-state VEPs are recorded to fast sequences of stimuli that are evenly spaced in time at intervals substantially shorter than the train of responses that follows a single stimulus. When recorded to checkerboard stimuli, the amplitude-versus-checksize characteristic is often non-monotonic and exhibits a so-called ‘‘notch’’ in many subjects. This could potentially be explained by constructive and destructive superposition, if one thinks of the steady-state VEP in terms of overlapping single-stimulus responses. We assessed how well amplitude-versus-checksize tuning can be predicted from

Doc Ophthalmol (2014) 129:13–55

17

simulated steady-state responses that are derived from transient responses. Methods It is methodologically challenging to obtain valid single-stimulus responses, because the longer inter-stimulus intervals (ISIs) that are typical for recording transient VEPs result, inter alia, in a different level of neural adaptation, compared to steady-state VEPs. We therefore estimated singlestimulus responses by deconvolving responses to fast m-sequences that had the same average inter-stimulus interval as the steady-state stimuli. Altogether, using checkerboard onset stimuli, we recorded three types of VEP responses. (1) Standard steady-state responses (short ISI). (2) Standard transient responses (long ISI). (3) Deconvolved m-sequence responses. Simulated steady-state responses were constructed from the responses (2) and (3) and compared to the original steady-state responses (1), both in the time domain and in the frequency domain. Five different check sizes ranging from 0.045 to 0.35 were tested in 20 subjects. Results In most subjects, steady-state responses that were constructed from long-ISI transient VEPs drastically exceeded the original steady-state responses in amplitude. There was a much closer match in amplitude when deconvolved m-sequence responses were used for construction. With respect to the amplitude-versus-checksize relationship, we found that in some subjects the simulated steady-state responses matched the original relationship better than did the single-stimulus responses. In other subjects, simulated steady-state responses were not a better match than the single-stimulus responses. Conclusions Assuming that the single-stimulus responses computed from the m-sequence recordings are representative of the steady-state single-stimulus response, the results suggests that constructive and destructive superposition of consecutive responses is one factor, but not the only one, that determines the amplitude-versus-checksize characteristic of the steady-state VEP. Funding Deutsche Forschungsgemeinschaft

Method: The 20 and 40 Hz descriptors were extracted from the DWT of the photopic ERG b-wave (flash intensity: 0.64 log cd.s./m2; background: 30 cd./m2) obtained from CSNB (N = 10) and CPCPA (N = 10) patients as well as from healthy subjects (N = 20). The ratio between these 2 descriptors (i.e. 40/20 ratio) was determined [minimal, maximal and mean values; standard deviations (SD) and coefficients of variation (CV)] in each group, and compared using Student t tests. A similar analysis was also conducted on ERGs obtained from RP patients (N = 20) of various etiologies. Results In controls, the 40/20 ratio is almost unity (mean ± SD: 1.04 ± 0.1; range 0.88–1.15; CV = 9.26 %). In comparison, it is significantly (p \ 0.0001) higher than normal (mean ± SD: 1.29 ± 0.2; range 1.14–1.59; CV = 15.5 %) in CSNB and significantly (p \ 0.0001) lower than normal (mean ± SD: 0.74 ± 0.1; range 0.61–0.82; CV = 9.4 %) in CPCPA due to a significantly reduced contribution of the 20 and 40 Hz, respectively. Results obtained from our RP patients were significantly more variable (mean ± SD: 0.77 ± 0.2; range 0.53–1.26; CV = 26 %). CPCPA-like ratios (i.e. 40/20 [ 1) were obtained on 12 RP patients, six had normal-like ratios (i.e. 40/20 & 1) and 2 CSNB-like ratios (i.e. 40/20 [ 1). Of interest, the latter categorisation of the RP disease process appears to remain throughout the degenerative process as revealed with long term follow-ups (30 years) of two CPCPA-like RP patients. Discussion: Our study suggests that the 20 and 40 Hz components identified with the DWT analysis of the photopic ERG b-wave reflects the respective contribution of the ON and OFF retinal pathways to the genesis of the b-wave. Although the exact retinal loci (probably bipolar cells) at the origin of these ERG components remain to be determined, results obtained from our RP cohort suggest that DWT analysis could be used to further segregate pathological responses that would otherwise appear functionally similar, irrespective of the stage reached in the degenerative process. Funding CIHR and Re´seau-Vision of the FRQ-S.

(11:30) Weighing ON–OFF pathway contribution to the photopic ERG with the discrete wavelet transform (DWT)

(11:45) Generating a non-naive ERG reference library using deterministic and stochastic models

M. Gauvin1, J.- M Lina2,3, R. K. Koenekoop1, J-M Little1, P. Lachapelle1 1 Department of Ophthalmology & NeurologyNeurosurgery, McGill University - Montre´al Children’s Hospital Research Institute, Montre´al, Que´bec, Canada; 2 De´partement de Ge´nie E´lectrique, E´cole de Technologie Supe´rieur, Montre´al, Que´bec, Canada; 3Centre de Recherches Mathe´matiques, Montre´al, Que´bec, Canada

Purpose Previous studies in our lab revealed that the b-wave of the photopic ERG is made of two major frequency components oscillating at *20 and *40 Hz, respectively. The purpose of this study was to determine the origin of these components by comparing frequency contents of photopic ERGs obtained from CSNB (ON pathway anomaly) and Congental Postreceptoral Cone Pathway Anomaly (CPCPA—an OFF pathway anomaly; see Lachapelle et al., Doc Ophthalmol 1998;95:35–54).

A. C. Fisher, R. Teymouri, A Eleuteri, R. P. Hagan Department of Medical Physics & Clinical Engineering, Royal Liverpool University Hospital; Department of Physics, University of Liverpool, Liverpool, UK Purpose ISCEV informs best clinical practice with Standards and Recommendations. However, the exact detail of data acquisition, data reduction and information extraction is dependent on the specific implementations of individual instrument manufacturers. The lack of universally-available reference datasets against which the clinical user can characterise their instrumentation and cross-reference their results with colleagues in other laboratories is a limitation to quality assurance, clinical governance and impedes cooperative work across institutions. Potentially, the mathematical construction of ‘deterministic’ artificial records based on templates agreed in consensus by experts is, in principle, trivial. Clearly, such simple noise-free data are of limited relevance as the highly challenging effects of electrical interference arising from the

123

18 EMG, blinks, eye-movements and mains supply sources are ignored. To be realistic and useful as reference sources, these effects must be modelled as ‘stochastic’ processes. Methods In 10 volunteers with no known eye disease, a range of ERG recordings was made over an extreme range of flash intensities under photopic conditions, explicitly not to ISCEV standards, over the range [1.0 … zero] cd per m in 10 log steps (Roland Ganzfeld). Data acquisition was with a bespoke amplifier with bandwidth [0.05 … 50000] Hz and sampling rate of 40 kHz. Electrodes were Liverpool Silver Thread DTLtype. In three sets of trials per subjects, optimal, blink- and eye movement- contaminated records were made. Data were reduced by 3rd-order IIR filtering ([0.1 … 100]Hz) and the ERG response recovered by selective averaging. Spontaneous noise artefacts (blinks and eye movements) were identified and characterised by Grubbs Test and robust (nonparametric) distribution-fitting based on median absolute deviation identification (Liverpool EDT-Eye ToolBox). The underlying continuous noise was modelled using an ARMA (autoregressive moving average) technique. A further series of records was synthesised using ISCEV ERG and ERG models as templates. All analyses were written in Matlab (Mathworks Inc.) and a click-and-go user interface (GUI) was implemented in MS Excel, internet-connected to the EDT-Eye MatSOAP server. Results A comprehensive library of non-naive ERG (clinically-based and synthetic) and PERG (synthetic) records was made. The GUI-interface freely-available as an MS Excel page over the Internet allows the user to produce a range of ERG record analogues with defined signal-to-noise ratios and artefact stochastic noise content. As an aside, the effect of user-selected pass-bandwidths on these records can be demonstrated with a webpage tool. Conclusion It is suggested that colleagues in ISCEV might consider this work as a contribution to extending its support for good scientific and clinical practice. The operation and accessibility of this library will be demonstrated in real-time over the Internet.

(12:00) Normal values of VEP amplitude, which measure: N75 to P100, or P100 to N170, or else? M. Bach Eye Center, Freiburg University, Freiburg, Germany Purpose One of the important and at the same time somewhat boring tasks for an electrophysiology lab is to obtain normal values. Even with well-standardized procedures, characteristic measures vary substantially among subjects, especially the amplitude. I will here concentrate on the VEP amplitude, where the ISCEV VEP Standard (Odom et al. 2010) recommends measuring ‘‘the amplitude of P100 from the preceding N75 peak’’. However, the N75 is not always clearly defined; Vaegan was especially adamant on this, ‘‘N75 is not a good zero reference level for P100 size measurement’’ (his symposium presentation 2005). Is it possible to find alternate amplitude measures that have less variability? Methods Following the ISCEV standard, we recorded VEPs to checkerboard reversal in 80 eyes of 40 normal subjects. Each condition (0.25 and 1 check size) was recorded twice. Amplitude measures were ‘‘NP’’ (N75–P100), ‘‘PN’’ (P100–

123

Doc Ophthalmol (2014) 129:13–55 N170), and ‘‘NPN’’ (mean of NP and PN). Test–retest reliability was quantified via the coefficient of variation (CV). Results In our sample, all peaks were clearly defined and could be determined automatically. N75 was deeper for 0.25 checksize compared to 1. The relative values of NP versus PN varied substantially between subjects. Mean ± SD of NP, PN, and NPN was 12.7 ± 6.0, 14.8 ± 6.3, and 13.7 ± 5.6 lV, respectively. The medians of CV and their 95 % confidence intervals were 0.062 [0.003–0.45], 0.071 [0.003–0.43] and 0.056 [0.005–0.37]. A permutation test (n = 106) confirmed the small, but significant difference between these measures (p = .037). Conclusions The NPN measure had 10 % less variability than NP; PN variability was highest. It is a trivial effort to measure PN in addition to NP. When both NP and PN amplitudes are well defined, we prefer the NPN measure due to its best test–retest reliability. When only NP or PN can be determined from pathologic traces, it is useful to use the corresponding normal values. (12:15) Normative sample size for reference intervals of visual electrophysiology parameters R. Hamilton Department of Clinical Physics, Royal Hospital for Sick Children and University of Glasgow, Glasgow UK Purpose ISCEV standards require the use of normative data, but do not provide recommendations on required sample size. Recommended normative sample sizes for physiological measurement parameters are usually small (B20) and presented with little statistical justification. In contrast, determination of population-based reference intervals in laboratory science, including recommended normative sample sizes, is the subject of an extensive literature including an approved guideline. The purpose of this study was to examine this literature to assess its applicability to visual electrophysiology parameters, to identify the distributions of commonly used visual electrophysiology parameters and therefore to draw conclusions about required normative sample sizes. Methods A literature search was conducted for relevant publications. Local normative datasets of PERGs (P50 and N95 amplitudes and peak times (PT), N = 54), of ERGs (aand b-wave amplitudes and PTs for light-adapted (N = 43) and dark-adapted (N = 48) ERGs) and of pattern-reversal VEPs (P100 amplitudes and PTs, 60’ (N = 85) and 12’ check widths) were tested for normality of distribution, applying logarithmic or Box-Cox transforms if necessary. The results were used to establish appropriate techniques for statistical treatment to determine reference intervals, and therefore, minimum recommended normative sample sizes. Results A two-tailed reference interval is defined by the 2.5th and 97.5th centiles, which represents the biological variation of the parameter of interest. Each of the two limits requires their 90 % confidence interval (CI) to be calculated, representing the measurement variation: the CI covers the true value of its reference limit with 90 % probability, and therefore indicates the reliability of the reference limit. A guideline ratio of 1:5 is recommended for measurement variation : biological variation, i.e. 90 % CI of a limit : reference interval. The majority of parameters had a normal distribution. The PERG P50 amplitude and the VEP P100 amplitude (12’ checkwidth) distributions normalised after logarithmic or Box-

Doc Ophthalmol (2014) 129:13–55 Cox transformations. Neither logarithmic nor Box-Cox transformation produced normal distributions for the ERG darkadapted b-wave amplitude, for the VEP P100 PT (60’ checkwidth) or for the ERG dark-adapted a-wave PT. For the normally-distributed parameters, the minimum normative dataset required to define the lower end of the 90 % CI of the lower reference limit using the so-called ‘robust method’ is 60 independent observations (one eye per subject). For the distribution-free parameters, the minimum number is 120. Conclusions To define adequately the limits of reference intervals with a specified (90 %) certainty for the ERG and VEP, normative datasets of at least 120 observations are required. Fewer may be necessary for the PERG, which shows normal distribution of its parameters. Such large normative datasets are unlikely to be achieved by individual laboratories, but rather by co-operative efforts across laboratories based on highly standardised methodology.

PA3: Glaucoma, retinal ganglion cells Monday, July 21, 16:15–18:00

(16:15) Assessing local retinal ganglion cell function in glaucoma using the multifocal electroretinogram L. Frishman1, L. Rajagopalan1, S. Viswanathan2, N. Patel1, R. Harwerth1 1

College of Optometry, University of Houston, Houston, TX; 2College of Optometry, State University of New York, New York, NY, USA Purpose The mfERG simultaneously measures local function at numerous locations in the retina. The photopic negative response of the mfERG (the mfPhNR) is thought to reflect local retinal ganglion cell function. In this study we examined the relation of mfPhNR amplitude to corresponding retinal structural measures using spectral domain Optical Coherence Tomography (sdOCT) and subjective visual field sensitivity (VS) using perimetry in nonhuman primates (NHP) with unilateral intraocular hypertension (OHT) as a model of experimental glaucoma. Methods mfERG recordings were made longitudinally in five macaque NHPs in up to 10 sessions, separated by 2–3 weeks for each subject. The visual stimulus display, 350 9 340, was composed of 19 black and white unstretched hexagons each subtending 7. The stimulus ‘‘flash’’ initially consisted of 1–6 bright frames, each occurring on 50 % of the frame changes (75 Hz), followed by 25 dark frames, repeating every 400 ms at the same location for 7 min. It was observed that mfPhNRs grew in amplitude for up to 5 frames, and then saturated; 5 frames were chosen for all subsequent testing. mfPhNR amplitudes (baseline to trough) were compared in experimental (Exp) eyes and fellow control (Con) eyes using a repeated measure ANOVA corrected with a post hoc Tukey test. The relationship between local mfPhNR amplitude and corresponding structural measures using sdOCT, i.e. sectoral retinal nerve fiber layer thickness (RNFLT) or retinal ganglion cell/inner plexiform thickness (RGC/IPLT), and with local VS measured with Humphrey Field Analyzer, were analyzed using linear regression analysis in individual subjects. Test–retest

19 variability for each functional and structural measure was calculated for Con eyes over all test sessions for each subject. Results Significant reductions in mfPhNR amplitudes were seen in glaucomatous (Exp) eyes relative to Con eyes after OHT was induced in NHPs (P \ 0.005), and the reductions increased over time. The mfPhNR amplitude in some regions of Exp retinas showed reductions that exceeded the test–retest variability of Con eyes in sessions before such changes in corresponding sectoral RNFLT, RGC/IPT or VS had occurred. Local mfPhNR amplitudes were well correlated with corresponding sectoral RNFLT in Con eyes (R [ 0.86, P \ 0.0001), due to the regional differences associated with eccentricity, and in Exp eyes as well, due both to regional differences and functional and structural loses (R [ 0.78, P \ 0.0001). Similar temporal effects were seen in relations between local mfPhNR amplitude and retinal ganglion cell/inner plexiform layer thickness. The changes in mfPhNR amplitudes observed in Exp eyes of NHPs, and correlations with structural measures and perimetric VS in NHPs were quite similar to those observed previously in human patients with open angle glaucoma in a study using similar stimulus conditions (Viswanathan S et al. IOVS 2010;51:ARVO E-Abstract 3262). Conclusions These findings indicate that the mfPhNR can be a useful tool for early detection and longitudinal assessment of functional changes in glaucomatous eyes.

(16:30) Multifocal electroretinography and regional choroidal blood flow in experimental glaucoma in rhesus macaques M. T. Nork, C. B. Y. Kim, K. M. Munsey, R. J. Dashek, J. N. Ver Hoeve Department of Ophthalmology and Visual Sciences, University of Wisconsin–Madison, Madison, WI USA Purpose To test a hypothesis that there is regional variation in the effect of experimental glaucoma on both retinal function and choroidal blood flow (ChBF). Methods Five rhesus macaques underwent unilateral laser trabecular meshwork destruction (LTD) to induce elevated intraocular pressure (IOP). IOPs were elevated for 56–57 weeks. Multifocal electroretinographic (mfERG) and multifocal visual evoked cortical potential (mfVEP) testing using a standard (215-1) m-sequence 241- and 7-element stimulus, respectively, were performed at regular intervals before and during the period of IOP elevation. At the time of sacrifice, the IOP was manometrically controlled at 35 (experimentally glaucomatous eye) and 15 (fellow control eye) mmHg. Five million 15-lm fluorescent microspheres were injected into the left ventricle. Regional ChBF was then calculated and plotted. Results Supranormal early waveform mfERG amplitudes and decreased late waveform first-order kernel (K1) amplitudes were found in all of the glaucomatous eyes. The supranormality was somewhat greater in the central macula. The K2.1 (kernel 2, slice1) root-mean-square (RMS) mfVEP response was inversely correlated (R2 = 0.97) with axonal loss. Total ChBF was reduced in the experimentally glaucomatous eyes. The mean blood flow was 893 ± 123 and 481 ± 37 ll/min in the control and glaucomatous eyes, respectively. Although

123

20 reduced everywhere, the ChBF showed regional variability with the greatest proportional decrement most often found in the central macula. Peripapillary ChBF decrease was proportional to the overall reduction in ChBF. Conclusions To our knowledge, this is the first demonstration of globally reduced ChBF in chronic experimental glaucoma in the non-human primate. Both the supranormality of the early mfERG waveforms and the reduction in ChBF tended to be greater in the macula, suggesting that there may be a causal link between ChBF and outer retinal effects in glaucoma.

(16:45) High frequency components of kernel 2 slice 1 (K2.1) mfERG following consecutive hemi-retinal axotomy and experimental glaucoma J. N. Ver Hoeve, C. B. Y. Kim, T. M. Nork Department of Ophthalmology and Visual Sciences, School of Medicine and Public Health, University of WisconsinMadison, Wisconsin, USA Purpose To compare the mfERG following selective ablation of retinal ganglion cell (RGC) axons with the effect of experimental glaucoma (ExpGl) within the same eye. Methods Three young adult male rhesus monkeys underwent placement of contiguous endodiathermy spots along the inferior portion of the optic nerve margin of one eye. Fundus photography and spectral domain OCTs (sdOCT) were obtained at baseline and at 9 months following axotomy. ExpGl was then induced by laser scarification of the trabecular meshwork in the same eye and intraocular pressure (IOP) elevations maintained for nearly 6 months. mfERGs were collected at baseline (n = 7), post hemi-retinal endodiathermy (n = 6), and post ExpGl (n = 9). The mfERG stimulus consisted of 241 equal-sized hexagons presented using a conventional fast m-sequence with a 75 Hz frame rate. mfERG data were collected and kernel 1 (K1) and K2.1 waveforms extracted using a VERIS 4.9 system (EDI, San Mateo CA). Waveforms were digitally filtered off-line to extract a low frequency component (LFC; \50 Hz) and a high frequency component (HFC; 70–140 Hz). Results In all three eyes, K2.1 HFC was the mfERG component most affected by axotomy. The amplitude of the K2.1 HFC was significantly reduced in retinal areas involving the inferior hemi-retinal axotomy while the response from the superior retina remained unchanged from baseline. The peak latency of the K2.1 HFC normally increases with distance of the stimulus patch from the optic nerve. This apparent latency shift also was absent in axotomized portions of the retina. Six months following induction of ExpGl in the same eye, mfERGs from the superior hemi-retina showed K2.1 HFC amplitude reduction and regularization of K2.1 HFC peak latency that is similar to the changes seen previously from the axotomized inferior retina. The residual K2.1 HFC from the axotomized retina was unchanged following ExpGl. Conclusions Recently, we have shown that hemi-retinal axotomy by endodiathermy produces complete loss of axons without disrupting outer retinal anatomy, opsin regulation, and retinal or choroidal blood flow (Dashek et al. IOVS 2013; 54: 3479–92). In that study, hemi-retinal axotomy selectively reduced both late (40–70 ms) K1 and total K2.1 root mean

123

Doc Ophthalmol (2014) 129:13–55 square amplitude. In the present study, we compare mfERGs from retinal areas that have sustained complete axon loss with the effects of chronic ExpGl in the same eye so that both areas of the retina were equally exposed to any generalized effects of IOP elevation or other experimental manipulations. We conclude that features of the high frequency oscillations contained in the second-order mfERG are spatially localized and specific to ganglion cell function. Funding Bright Focus Foundation, P30 EY016665, Research to Prevent Blindness, Inc., and The Wisconsin National Primate Research Center P51RR000167/P51OD011106

(17:00) The relationship among N2 amplitudes of multifocal ERGs, retinal sensitivity and retinal nerve fiber layer in glaucomatous eyes and normal eyes F. Kato, G. Miura, S. Shirato, S. Yamamoto Department of Ophthalmology and Visual Science, Chiba University Graduate School of Medicine, Chiba, Japan Purpose To investigate whether or not a significant correlation exists among the amplitude of the multifocal photopic negative response (mfPhNR), retinal sensitivity and retinal nerve fiber layer thickness (RNFLT) in glaucomatous eyes and normal eyes. Methods Thirty-one glaucomatous eyes and 10 normal eyes were studied. The mfPhNR was recorded to red stimuli presented on a blue background. Responses from the central five elements within 20 degrees were analyzed. The amplitudes of the negative trough were compared with retinal sensitivity measured by Humphrey Field Analyzer (HFA) and RNFLT by spectral-domain optical coherence tomography (OCT) of the corresponding area. Results The PhNR amplitude was significantly reduced in the glaucoma group in all areas as compared with the normal group (p = 0.02 for superior retina; p = 0.0007 for central retina; p = 0.002 for inferior retina). There was a significant correlation between the PhNR amplitude and the mean deviation (MD) of the HFA for central and inferior retinas (r = 0.48, p = 0.001; r = 0.42, p = 0.005, respectively) and a trend for superior retina (r = 0.41, p = 0.08). There was a significant correlation between the PhNR amplitude and RNFLT (r = 0.55, p = 0.0002 for superior retina; r = 0.67, P \ 0.0001 for central retina; r = 0.41, p = 0.01 for inferior retina). Conclusions The multifocal PhNR response recorded to these stimuli may receive some contributions from the retinal ganglion cells, and the mfPhNR amplitude can be used as an objective measurement of the retinal ganglion cell function in glaucoma patients.

(17:15) Impact of different OCT brands on structure– function analysis: a pilot study L. Brandao1, A. A. Ledolter1,2, A. Scho¨tzau1, A. M. Palmowski-Wolfe1 Department of Ophthalmology, University of Basel, Basel Switzerland; 2Medical University of Vienna, Department of Ophthalmology, Vienna, Austria

Doc Ophthalmol (2014) 129:13–55 Purpose The brands of spectral OCT used, and their respective segmentation programs, might influence structure function analysis results. In this regard, we assessed retinal full macular thickness (MT) and ganglion cell layer-inner plexiform layer (GCIPL) thickness using the CirrusTM and the SpectralisTM Optical Coherence Tomography (OCT) in primary open angle glaucoma (POAG) patients and controls. Findings from Cirrus and Spectralis were compared to automated perimetry (AP) and 2F-mfERG. Methods 23 eyes were included: 10 POAG and 13 controls. OCT: fast macular thickness protocol, both CirrusTM and SpectralisTM. AP: Octopus 101, G2. 2F- mfERG: VERIS 6.06TM, FMSIII; recording parameters: bandpass filter (BPF): 1–300 Hz, 103 Hexagons, M-sequence 213–1: LMax 100 cd/m 2 , Lmin \ 1 cd/m2, global flash:200 cd/m2. Analysis: AP: mean defect (MD in dB); 2F-mfERG: Responses filtered at 1–200 Hz. RMS was calculated for the focal flash response at 15–45 ms (DC) and for the global flash responses at 45–75 ms (IC1) and 75–105 ms (IC2). OCT and mfERG values from the central 10 were converted into logarithmic scale for comparison to AP. OCT: MT and GCA (GCL + IPL) as per Cirrus Software and Heyex Segmentation Software beta-version. For Spectralis, segmentation of GCIPL was also corrected manually. Results Bland–Altman analysis a showed significant agreement between OCTs in MT (p \ 0.0001, IC 95 % 12.4–23.9, SD ± 2.9). However, analysis of the GCIPL was only significant when Spectralis was compared to Cirrus Minimum GCIPL values (p = 0.0039, IC 95 % -6.3to13, SD ± 4.9). Mean MT values obtained in Cirrus and Spectralis were: 296.5 lm (SD ± 17.3) versus 313.9 lm (SD ± 16.2) for patients and 306.3 lm (SD ± 12.5) versus 325.2 lm (SD ± 10.6) for controls. Mean GCIPL thickness in Cirrus and Spectralis were: 67.2 lm (SD ± 10.4) versus 64.2 lm (SD ± 13.3) in patients and 80.1 lm (SD ± 4.95) versus 81.3 lm (SD ± 3.9) in controls. Mean Minimum GCIPL thickness in Cirrus were: 59.6 lm (SD ± 11) in patients versus 79 lm (SD ± 4.8) in controls. Glaucoma differed significantly from Control in MT and GCIPL when measured with Spectralis, but only in GCIPL when measured with Cirrus. LogMT and logGCIPL had a negative correlation to MD in both OCTs, especially in POAG where this was highly significant. While logarithmic mfERG values of the induced components differed significantly between glaucoma and control, there was no significant correlation to either measure in Cirrus or Spectralis. Conclusions While MT and GCIPL minimum values of Cirrus agreed significantly with Spectralis values, these values were not interchangeable, as for instance MT was always higher in Spectralis. Interestingly, MT was especially higher in the control group leading to the observation that MT of controls differed from glaucoma only with Spectralis but not Cirrus. In regard to structure–function analysis when either OCT was compared to 2F-mfERG or AP, results were comparable. Acknowledgments AMP: SNF (Swiss National Science Foundation) NMS 1823, LHW Stiftung Lichtenstein. Thank you for providing the segmentation analysis: Carl Zeiss Meditec, Feldbach, Switzerland, esp. Matthias Monhart; Heidelberg Engineering GmbH, esp. Diana Helling and Sebastian Lukas.

21 (17:30) Differences in functional loss associated with ganglion cell complex thinning between patients with glaucoma and patients with postoperative macular hole S. Machida, K. Tamada, T. Ohzeki, Y. Gotoh, D. Kurosaka Department of Ophthalmology, Iwate Medical University, Iwate, Japan Purpose To compare the changes in the photopic negative response (PhNR) of the focal macular ERG (fmERG) caused by a thinning of the ganglion cell complex (GCC) between patients with open angle glaucoma (OAG) and after successful macular hole (MH) surgery. Methods Twenty-seven eyes of 27 OAG patients (OAG group), 28 eyes of 28 MH patients (MH group), and 23 eyes of 23 normal volunteers (control group) were studied. The OAG patients had early stage glaucoma. The MH patients had been successfully treated with vitrectomy 12 months earlier. The averaged GCC thickness in the macular area (15 degree in diameter) was measured by spectral domain OCT. The fmERGs were elicited by a 15 stimulus centered on the fovea. Results The GCC was significantly thinner in the OAG and MH groups than in the control group (P \ 0.0001). The PhNR amplitude and PhNR/b-wave amplitude ratio were significantly reduced in the OAG group compared to the control (P \ 0.0001), while no significant reduction of the PhNR amplitude was observed in the MH group. The PhNR amplitude and PhNR/b-wave amplitude ratio were significantly correlated with the GCC thickness in the OAG group (R2 = 0.741 for the PhNR amplitude, R2 = 0.564 for the PhNR/bwave amplitude ratio, P \ 0.0001). There were significant differences between the OAG and MH groups in the slopes of regression lines plotting the GCC thickness and the PhNR amplitude (P \ 0.05) and the PhNR/b-wave amplitude ratio (P \ 0.0005). Conclusions The thinning of the GCC affects retinal ganglion cell function in the macula area differently for patients with OAG and patients after macular hole closure. Funding JSPS KAKENHI grant No. 24592677

(17:45) Biphasic PERG decrements correlate with timedependent changes in spontaneous and light-evoked retinal ganglion cell activity in a mouse model of blastinduced traumatic brain injury M. Harper1, L. M. Dutca1,2, F. R. Blodi3, M. Shankar3, M. Anderson1,4, A Hedberg-Buenz4, R. H. Kardon1,2, S. F. Stasheff3 1

Center of Excellence for Prevention and Treatment of Visual Loss, Iowa City Veterans Administration Medical Center; Departments of 2Ophthalmology and Visual Science, 3Pediatrics (Neurology), 4Molecular Physiology and Biophysics, The University of Iowa, Iowa City, IA, USA Purpose To analyze the in vivo and in vitro function of retinal ganglion cells (RGCs) after blast-mediated traumatic brain injury (TBI).

123

22 Methods Mice were exposed to an overpressure wave (20 PSI) directed to the head using a custom-built blast chamber. Analysis of in vivo RGC structure and function was performed using optical coherence tomography (OCT) and PERG at 7 days, 5 weeks and 4 months after blast exposure. Individual RGCs (*250 cells for each time point) from freshly dissected whole-mounted retinas were monitored using a multielectrode array (MEA) at identical time points. The spontaneous activity and the light evoked responses (to full-field flashes) for each RGC were measured and compared with retinas from control mice and across time-points. The statistical analysis was performed using the Mann–Whitney test. The dendritic arborization of individual RGCs was analysed in vitro using GFP labelled RGCs. Results PERG amplitudes were decreased 7 days following blast exposure compared to baseline amplitudes. The PERG response recovered to baseline amplitudes 5 weeks post injury. Subsequent PERG analysis 4 months post blast showed significantly decreased PERG responses compared to baseline. Decreased thickness of the RGC complex layer was observed by OCT at all time points. Seven days after blast exposure, the spontaneous activity of RGCs was slightly increased compared to controls. The fraction of cells that responded to the onset of light decreased significantly, while the median response amplitude (spikes/s) and response duration were similar to those of control RGCs. The median amplitude of responses to light offset increased significantly, while response duration decreased. Five weeks post-blast, spontaneous activity was strikingly increased compared to both controls and the 7 day recordings. The fraction of cells responding to the onset of light was similar to that at day 7, while median response amplitude and response duration were significantly increased compared to both control and 7 day post-blast. OFF response amplitude was significantly increased, while response duration was similar to control levels. By 4 months post-blast, all the measures of RGC physiology had recovered to near-normal values and were similar to those at 7 days post-blast. Cell counts of the entire retina and dendritic analysis revealed significantly decreased RGC number and abnormal dendritic fields, which correlated to abnormal RGC receptive fields observed in the MEA analysis. Conclusions Exposure to blast induces dramatic alterations in the physiological activity of RGCs after an initial period of relatively normal function. These latent effects may indicate an optimal time interval during which such dysfunction may be prevented or ameliorated. The return to more normal RGC function by later time points may reflect the physiology of subpopulations that survive longterm, after a substantial proportion of RGCs die, reflected by the significant decrease in PERG response and abnormal dendritic fields. The differential effects of blast injury on ON and OFF responses at various times following exposure may indicate differential susceptibility of particular RGC types to this injury. Better understanding of the RGC physiology after blast exposure will help in the development of improved clinical testing and treatment of those suffering from TBI.

123

Doc Ophthalmol (2014) 129:13–55

PA4: animal models Tuesday, July 22, 9:00–10:30 am

(9:00) Retinal dysfunction in the GPR179 form of complete congenital stationary night blindness N. Peachey Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA Purpose Complete congenital stationary night blindness (cCSNB) is a genetically heterogeneous condition, for which five genes have been implicated (NYX, GRM6, TRPM1, GPR179, LRIT3). This study focuses on the GPR179 form. Methods Retinal electrophysiology was evaluated by ERG studies of anesthetized animals. Protein localization was evaluated by immunohistochemistry. Results GPR179 was identified based on the ERG phenotype of the nob5 mouse, which has a dramatic reduction in b-wave amplitude under dark- and light-adapted conditions. Unlike other mouse models of cCSNB, Gpr179-nob5 mice retain a small ERG b-wave that is readily seen when low luminance flashes are presented to the dark-adapted mouse. This component is enhanced when flash duration is extended, indicating that mGluR6 is still able to activate TRPM1 channels. A similar response is obtained from mice lacking both RGS7 and RGS11, which are both abnormally localized in the Gpr179nob5 retina. Shortly after the original description of the Gpr179-nob5 mouse, a European group identified the same allele in a transgenic colony. Conclusions The GPR179 form of cCSNB has a distinct phenotype from other subtypes of this condition. It will be important to determine if comparable phenotypic variability is found in human patients. The presence of the Gpr179-nob5 allele in multiple mouse colonies indicates that it may be widespread as a background allele.

(9:15) Correction of the dark adaptation defect in RLBP12/2 mice is a model to assess viral vectors for RLBP1-associated retinal dystrophy gene therapy C. Bigelow1, S. M. Hanks1, J. Vrouvlianis1, O. Turner2, G. Argentieri2, C. Hayden2, G. Bounoutas1, V. W. Choi1, T. L. Mc Gee1, A Gujar1, J. Aranda1, H. Li1, T. P. Dryja1, S. R. Police1, B. D. Jaffee1 1 Novartis Institutes for BioMedical Research, Cambridge, M. A. US; 2Novartis Institutes for BioMedical Research, East Hanover, NJ, USA

Purpose RLBP1-associated retinal degeneration is caused by absent or non-functional cellular retinaldehyde binding protein (CRALBP) arising from recessive mutations in the RLBP1 gene. Patients exhibit very slow dark adaptation from an early age and can be legally blind by middle age due to associated photoreceptor degeneration. The purpose of this study was to develop and characterize a mouse model to assess the ability of recombinant adeno-associated (rAAV) viral vectors to repair slow dark adaptation observed in RLBP1-/- mice.

Doc Ophthalmol (2014) 129:13–55 Methods The effect of age on retinal morphology was assessed in 4–16 month old RLBP1+/+ and RLBP1-/- mice. Evaluations relied on viewing sections stained with hematoxylin and eosin or via image processing to measure retina and outer nuclear layer thickness. Dark-adapted ERGs were compared between RLBP1+/+, RLBP1±, and RLBP1-/- mice using a 15-step intensity response series (-1.0 to 3.7 log scot cd s m2). ERG a-wave amplitude recovery after a photobleach (16 xenon flashes: 3.7 log scot cd s m2) was used to assess dark adaptation rate. 3.6 9 108 vg/eye of AAV-pRLBP1-hRLBP1 vectors (NVS1 or NVS2) with different serotypes (AAV2 or AAV8) were injected subretinally in RLBP1-/- mice to test the ability of gene delivery to improve the dark adaptation rate. Dark adaptation based on a-wave amplitude recovery postbleach was quantified longitudinally for 2–49 weeks postinjection. Results Retinal morphology did not differ notably up to 16 months of age or between genotypes based on light microscopy examination or image processing. No genotypedependent differences in dark-adapted a-waves were detected. However, slow dark adaptation was observed in RLBP1-/mice compared to their corresponding wild-type or heterozygous controls with *3 h post-bleach a-wave recoveries of 14, 89, and 97 %, respectively. Both NVS1 and NVS2 significantly increased the rate of dark adaptation compared to naı¨ve RLBP1-/- mice. NVS2 significantly improved dark adaptation 2–49 weeks post injection (p \ 0.05 to p \ 0.0001) while NVS1 resulted in a significant effect for 12–49 weeks post injection (p \ 0.05 to p \ 0.01). Serotype-related efficacy differences were significant: eyes that received NVS2 (AAV8) exhibited a statistically significant 1.5- to 3.7-fold increase in a-wave amplitude versus those receiving NVS1 (AAV2) for at least 49 weeks post-injection (p \ 0.05 to p \ 0.0001). Conclusions Although RLBP1-/- mice do not exhibit retinal degeneration, they have very slow dark adaptation similar to that observed in patients with RLBP1-associated retinal dystrophy. ERG-based visual function assessments of RLBP1-/- mice indicate that their slow dark adaptation can be corrected with viral vector-mediated gene delivery and that the model possesses sufficient sensitivity to differentiate viral vectors with different properties.

(9:30) mGluR5 is an important molecule mediating Mu¨ller cell contribution to the rat’s flash electroretinogram Z. Q. Yin, J. Dai, Z. Tao, Y. Chen, Y. Li Department of Ophthalmology, Southwest Hospital, Third Military Medical University Country, Chongqing, China Purpose To investigate the mechanisms of the Mu¨ller cells contribution to the light-evoked ERG responses by regulating the ion-water homeostasis and recycling of neurotransmitters. Methods First, a rat Mu¨ller cell inhibition model was established with subretinal space injection of DL-alphaaminoadipic acid (DL-AAA). Immunofluorescence staining, qualitative real-time polymerase chain reaction (qRT-PCR), western blotting, Calcium imaging and patch clamp techniques were further utilized to clarify the changes of different molecules and ion concentrations.

23 Results We found an obvious trend of inverse correlation between the activation of metabotropic glutamate receptor 5 (mGluR5) expressed by Mu¨ller cells and ERG b-wave amplitude, especially the descending limb of the ERG b-wave. In spite of suppressing the optic nerve (ON) bipolar cells responses blocked by mGluR6 agonist, the amplitude of the b-wave could be rescued by inhibition of mGluR5. Then the mGluR5 downstream Gaq-Ca2+ axis was activated when mGluR5 was upregulated with administration of DL-AAA or its activator. Meanwhile, activation of mGluR5 downregulated the expression of Aqp4 and Kir4.1, and also suppressed the inward rectified potassium current because we confirmed that they formed a macromolecular complex. Conclusions The sum change of electrical signals derived from [Ca2+]i and [K+] may be the molecular basis of the Mu¨ller cells contribution to the ERG components. Our study provides evidence mGluR5 mediates the contribution of Mu¨ller cells to the light-evoked responses.

(9:45) Delayed photoreceptor degeneration in neonatal hypoxic-ischemic retinopathy (NHIR) S. Jung1, P. Wintermark1,3, Pierre Lachapelle2,3 1

Department of Pediatrics, Division of Newborn Medicine, Montreal Children’s Hospital-Research Institute, McGill University; 2Department of Ophthalmology, Montreal Children’s Hospital-Research Institute, McGill University; 3 Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada Purpose It has previously been demonstrated that neonatal hypoxia–ischemia (HI) induces irreversible damage to the structure and function of the inner retina while sparing the outer retina until postnatal day 60 (P60) in rats. The purpose of this study was to determine whether the preservation of photoreceptors persisted over time. Methods Neonatal HI was induced in 10-day-old male LongEvans rat pups by left common carotid artery ligation followed by 2-h hypoxia (8 % oxygen). Sham operated rats served as control. Flash ERGs were recorded from the left and right eyes at P60, 90, 120, 150, 200, 250, 300 and 350, after which the eyes were removed for retinal histology. Comparisons were made between the control group and the HI group, and between P60 and later time points. Results As previously shown, there was no significant difference in the scotopic a-wave amplitude between groups at P60 while the scotopic and photopic b-wave amplitudes were significantly attenuated (to 30.75 and 14.05 % of control, respectively) in the left eye (i.e., ipsilateral to the carotid ligation) of the HI animals. Interestingly, however, the left eye of the HI animals also showed a gradual decrease in the a-wave amplitude over time (significant at P120 and onwards compared to P60) such that they were significantly lower compared to control from P90 and onwards (70.54, 57.21, 55.41, 50.33, 49.52, 44.93, and 36.00 % of control at respective ages). In contrast, the right eye (i.e., contralateral to the carotid ligation) of HI animals and both eyes of control animals did not show changes in the a-wave amplitude over time compared to their respective P60 values. The amplitudes of scotopic and photopic b-waves did not change significantly from P60 in all groups except in the right eye of the HI group, which

123

24 showed a significant increase in the amplitude of photopic b-wave at P200 and P250 compared to P60. In line with the functional findings, retinal histology from the left eye of HI animals showed thinning of the outer nuclear layer (ONL) at P350 compared to P60, while no such differences were observed in the inner retinal layers. Interestingly, retinal thinning also showed a retinotopic distribution where the extent of injury was most severe in the center and far periphery and relatively milder in the mid-periphery. Conclusions Our results suggest that neonatal HI causes a different time course of outer and inner retinal damage in rats. That is, until P60 it is the inner retinal structure and function that are impaired by an HI insult while no apparent deficits are observed in the photoreceptors. However, beyond P60 the photoreceptors undergo significant functional and structural degeneration, whereas the inner retinal injury does not progress. Despite the difference in time course, both outer and inner retinas are most affected at the center and periphery of the retina, pointing to both common and differential potential mechanisms underlying HI-induced retinal injury. Funding CIHR, FRQS, SickKids Foundation, Re´seau Vision, RI-MUHC/MCH, Foundation of Stars

(10:00) Origin of intrinsic optical signals in the retina A. Naderian, L. Bussieres, S. Thomas, F. Lesage, C. Casanova University of Montreal, Montreal, Canada Purpose Optical imaging of retinal intrinsic signals (RIS) is a relatively new method that measures changes in light reflectance of the retina that occur following visual stimulation. The exact physiological mechanisms behind RIS remain to be determined. Thus, the aim of this study was to examine the characteristics of RIS as a function of stimulation parameters and to determine the cellular origin of these signals. Methods Experiments were performed on anesthetized rabbits under scotopic conditions. Optical imaging of retinal intrinsic signals (RIS) was performed using a fundus camera capturing the reflected light from the retina, illuminated by near infrared light (750 nm). ERGs were concurrently recorded. The activity-dependency of RIS was evaluated by observing the effects of stimulus intensity on intrinsic responses amplitude and by comparing the RIS responses with the ERG a- and b-waves. Retinal stimulation consisted of a diffuse flash of green light (546 nm). The cellular origin of RIS was studied by blocking different retinal cell types by injecting various pharmacological agents intravitreously (aspartate, TTX, APB and PDA). Results Strong correlations were found between RIS amplitude and stimulus intensity (r2 = 0.879) as well as with the ERG b-wave amplitude (r2 = 0.977). No RISs were observed when the photoreceptor activity was isolated with aspartate, suggesting that photoreceptors are not the origin of RIS. Following the injection of TTX, a voltage-sensitive sodium channel antagonist, a small but significant decrease (20 %, n = 16) in RIS amplitude was observed, suggesting a partial role for ganglion cell activity in intrinsic signals. Following TTX injections, the administration of bipolar cell blockers (PDA and APB) affected the RIS amplitude in a concentration-dependent

123

Doc Ophthalmol (2014) 129:13–55 manner (20 % decrease for 2 mM of APB and a total disappearance for 10 mM). Conclusions Our results indicate that RIS are dynamic signals that can reflect retinal activity levels. In rabbits, optical signals originate from the inner retina with a strong contribution of bipolar cells; photoreceptors do not contribute to the genesis of RIS. Supported by a NSERC grant.

(10:15) Short and long exposures: same consequences in juvenile Light-Induced Retinopathy (LIR)? A. Polosa, W. A. Igrine, L-A Chevrolat, H. Bessaklia, P. Lachapelle Department of Ophthalmology & Neurology/Neurosurgery, McGill University/Monte´al Children’s Hospital Research Institute, Montre´al, QC, Canada Purpose Juvenile rats are more resistant to LIR than adult ones. While in adult rats, a 6 day long exposure to a light intensity of 10,000 lux will completely destroy the photoreceptor layer, Juvenile rats exposed for 14 days (between P1428) to the same light stress will present with a significantly better preserved retina. However, with time, the retinopathy of the Juvenile rats continue to degenerate to finally yield a LIR comparable to that observed in adult rats immediately following the cessation of light exposure. The purpose of this study was to determine, in Juvenile rats, the minimal duration of light-exposure required to cause, with time, an adult rat-like LIR. Methods At postnatal (P) day 14, neonatal Sprague–Dawley (SD) albino rats were exposed to a bright cyclic light (10 k lux, 12 h dark/12 h light) for various lengths of time [1, 2, 3, 6, 9, 12 and 14 consecutive days (d)]. Photopic and scotopic ERGs and retinal histology were performed at each time point. Animals exposed for 1 and 14 days were also re-evaluated at P37 and P70. Results A dose-dependent decrease of all ERG components is observed immediately following the cessation of light exposure (i.e. 1 day is less damaged than 14 days of exposure), with the scotopic ERG showing more attenuation than the photopic ERG [scotopic ERG was 15 % and photopic ERG was 25 % of controls, p \ 0.05]. When animals were retested at P37 and P70, only the a-wave of the scotopic ERG showed significant differences with age-matched 14 day exposed animals [a-wave at P70: 1 day: 88 % and 14 day: 32 % of controls, p \ 0.05]. Similarly, with increasing duration of exposure, a progressive loss of the photoreceptor outer nuclear layer (ONL) occurred. The first signs of photoreceptor cell death and typical hemiretinal asymmetry were noticed with less than 2 days of exposure. Interestingly, while no obvious damage could be documented following 1 day, when re-examined at P70, photoreceptor loss was almost identical to the damage produced by 14 days of light exposure [ONL thickness at P70: 1 day = 33 % vs. 14 day 29 % of control, p [ 0.05 (values taken at 1,000 lm from the optic nerve head)]. Conclusions Electrophysiological and histological findings suggest a dose-dependent retinopathy when the retina is assessed immediately following the cessation of exposure. However, our long-term studies also reveal that one day of

Doc Ophthalmol (2014) 129:13–55 light exposure is sufficient to produce the same LIR degeneration as that resulting from a 14 day long exposure, but the rate of photoreceptor loss appears being steeper with shorter than longer light exposure durations. One possible explanation could be that brief exposures are too short to induce either sufficient amounts of neurotrophic factors or antioxidants to efficiently counteract the damaging effect of bright light exposure. Further studies are needed to confirm this hypothesis. Funding NSERC.

PA5: Inherited retinal diseases Tuesday, July 22, 11:00 am–12:30 pm

(11:00) Atypical RPE65 related retinal dystrophy G. Holder, S. Hull, A. G. Robson, R. Mukherjee, A. T. Moore, A. R. Webster Moorfields Eye Hospital; UCL Institute of Ophthalmology, London, UK Purpose Mutation in RPE65 is usually associated with autosomal recessive Leber Congenital Amaurosis. The present study reports six cases of atypical RPE65 related retinal dystrophy. Methods Patients were identified from the genetics database at Moorfields Eye Hospital. Most patients had received structural imaging and electrophysiological assessment of function. Results Two patients with hypomorphic mutations had a rodcone dystrophy phenotype on ERG. Both presented with night blindness aged 2–3 years but maintained useful visual acuity into their 20’s. One patient had a fundus albipunctatus-type appearance, with some ERG recovery on extended dark adaptation. Three patients had dominantly inherited disease; one presented with nyctalopia in late teens and reduced central vision early 30’s; the other two (cousins) had onset between 35 and 45 years of age. Conclusions Atypical homozygous RPE65 retinal dystrophy includes mild disease and fundus albipunctatus-like appearance. Dominant disease can occur with variable severity and phenotype.

(11:15) Phenotypic variability in RDH12 retinopathy R. Sui1, X. Zhou1, H. Li1, X. Wang2, X. Liang1, R. Chen2, F. Dong1 1

Peking Union Medical College Hospital, Beijing, China; Baylor College of Medicine, Houston, TX, USA

2

Purpose To describe phenotypic variability and report genetic defects in a cohort of Chinese patients with mutations in RDH12 gene. Methods Twenty patients from 18 unrelated families with pathogenic RDH12 mutations on both alleles were studied. Detailed ophthalmologic examination, fundus photography, spectral domain optical coherence tomography (SD-OCT), and electroretinograms were performed. The coding exons and

25 intron–exon boundaries of RDH12 were analyzed by Sanger sequencing. Results Twenty different mutations in RDH12 were detected in the 18 families in the study, with 11 (55 %) of these changes being novel. Among the patients, 9 probands were diagnosed as severe early childhood onset retinal dystrophy (SECORD), 4 probands as Leber congenital amaurosis (LCA), and 5 probands as autosomal recessive retinitis pigmentosa (ARRP). Characteristic RDH12 retinal phenotypes included: type 1, the petal-like macular coloboma with surrounding pigment changes; type 2, wide-spread pigment changes, RPE atrophy and ‘‘gold foil refection’’ maculopathy; type 3, heavy and confluent pigment proliferation involving the macular region. The OCT showed macular depression and chorioretinal atrophy without normal laminar structure. One patient had a subretinal hemorrhage. ERG recordings ranged from non-recordable in LCA to mildly decreased in ARRP. Novel missense mutation p.V146D is the most frequent variant, accounting for 22 % of the genotypes. Eighty percent of ARRP patients carried missense or non-frameshift changes in both alleles, while at least one loss of function allele was present in nearly 50 % of early onset retinal dystrophy cases. Conclusions Wide but characteristic phenotypes were observed in RDH12 retinopathy. RDH12 mutation is the common cause of early onset retinal dystrophy. Loss of function allele is more frequent in SECORD and LCA than that of ARRP.

(11:30) Clinical and genetic characteristics of childhoodonset Stargardt disease K. Fujinami125, J. Zernant6, A. G. Robson4, 5, K. Tsunoda1, Y. Ozawa3, K. Tsubota2, R. Allikmets6, 7, M. Michaelides4, 5, A. T. Moore4, 5, G. E. Holder4, 5 1 Laboratory of Visual Physiology, National Institute of Sensory Organs, National Hospital Organization, Tokyo Medical Center, 2-5-1, Higashigaoka, Meguro-ku, Tokyo 152-8902, Japan; 2Department of Ophthalmology, Keio University, School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan;3Laboratory of Retinal Cell Biology, Department of Ophthalmology, Keio University, School of Medicine, 35 Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan; 4 UCL Institute of Ophthalmology, 11-43 Bath Street, London, EC1 V 9EL, UK; 5Moorfields Eye Hospital, City Road, London, EC1 V 2PD, UK; 6Department of Ophthalmology, Columbia University, 630 West 168th Street, New York, NY, 10032, USA; 7Department of Pathology and Cell Biology, Columbia University, 630 West 168th Street, New York, NY,10032, USA. Purpose To describe the clinical and molecular characteristics of 42 patients with childhood-onset Stargardt Disease (STGD). Methods The patients were retrospectively ascertained from the clinical database. A detailed history and the results of ophthalmic examination were available. Investigations included colour fundus photography, spectral domain optical coherence tomography (SD-OCT), and pattern and full-field electroretinography (PERG, ERG) that incorporated the ISCEV standards. The ERGs were used to classify patients

123

26 into three groups (Lois et al. 1999); dysfunction confined to the macula (group 1), macular and generalised cone dysfunction (group 2) and macular and generalised cone and rod dysfunction (group 3). The entire coding region and splice sites of ABCA4 were screened using a next generation sequencing-based strategy (Zernant et al. 2011). The molecular genetics and electrophysiological findings of the childhood-onset STGD patients were compared to those of a group of adult-onset patients (N = 64). Results The median ages of onset and the median age at baseline examination were 8.5 (range 3–16) and 12.0 years (range 7–16), respectively. The median baseline LogMAR visual acuity was 0.74. At baseline, 26 (67 %) of 39 patients with available photographs had macular atrophy with macular/ peripheral flecks; eleven (28 %) had macular atrophy without flecks; one (2.5 %) had numerous flecks without macular atrophy; and one (2.5 %) had a normal fundus appearance. Flecks were not identified at baseline in 12 patients (31 %). SD-OCT detected foveal outer retinal disruption, to a variable degree, in all 21 patients with available images. Electrophysiological assessment demonstrated group 1 type ERG in 9 patients (36 %), Group 2 type ERG in 1 patient (4 %), and Group 3 type ERG in 15 patients (60 %). At least one diseasecausing ABCA4 variant was identified in 38 patients (90 %), including 13 novel variants; two or more variants were identified in 34 patients (81 %). Compared to patients with adult-onset STGD, patients with childhood-onset STGD more frequently harboured two null variants (5 % compared to 18 %). A higher proportion of patients with childhood-onset STGD were in ERG group 3 (60 %) compared to adult-onset STGD (31 %). Conclusions Childhood-onset STGD is associated with severe visual loss, early morphological changes, and often generalized retinal dysfunction despite often less severe clinically apparent fundus abnormalities. The relatively high proportion of null ABCA4 variants, and the high percentage of Group 3 disease, suggest that earlier onset disease is generally caused by more deleterious variants than those present in adult-onset disease. (11:45) Abnormal post-photoreceptor connectivity and thickening of the outer nuclear layer in enhanced S-cone syndrome M. Sustar1, D. Perovsek1, I. Cima2, B. Stirn1, M. Hawlina1, J. Brecelj1 1 University Medical Centre, Eye Hospital, Ljubljana, Slovenia; 2University Eye Clinic, University Hospital, Sveti Duh, Zagreb, Croatia

Purpose To correlate functional ERG abnormalities of enhanced S-cone syndrome (ESCS) with structural findings and retinal architecture, as obtained by spectral domain optical coherence tomography (SD-OCT). Methods Four patients with genetically confirmed ESCS underwent standard full-field and multifocal ERGs, as well as extended S-cone and ON/OFF ERG protocols. SD-OCT of the macula was also carried out and longitudinal reflectivity profiles (LRPs) were calculated. Results All four patients exhibited characteristic full-field ERG findings for ESCS, with delayed responses with similar

123

Doc Ophthalmol (2014) 129:13–55 waveforms under both scotopic and photopic conditions to the same intensity flash. The full-field ERG amplitudes showed considerable variation between the patients and were not related to their clinical appearance or the extent of the visualfield defects. Multifocal ERG showed correlation of central retinal function with visual acuity. The ERG to S-cone specific stimulation revealed predominant activity of the S-cone system in all four patients. The ON/OFF ERG recording revealed both ON-response and OFF-response activities in three of the patients, while in the remaining patient, only OFF-response activity was recognized. The LRP profiles of the SD-OCT showed significantly thickened outer nuclear layer in all four patients. In the patient with absent ON-response activity, the LRP showed reduced number of hyper-reflectivity peaks at the inner plexiform layer (IPL). Similar lack of lamination peaks at the IPL was also detected in our patients with complete type of congenital stationary night blindness, in which ON-pathway dysfunction is a characteristic ERG finding. Conclusion Patients with ESCS exhibited typical full-field ERG waveform abnormality, S-cone hypersensitivity and thickening of the outer nuclear layer on the OCT. In addition, they showed abnormal retinal post-photoreceptor connectivity, through the S-cone-related OFF-bipolar cell activity, that might be associated with structural alterations at the level of the inner plexiform layer.

(12:00) FHONDA syndrome: foveal hypoplasia and optic nerve misrouting without albinism caused by recessive mutations in SLC38A8 M van Genderen, J. Poulter, M. Al-Araimi, I. Conte, C. F. Ingleheam, C. Toomes Leeds Institute of Molecular Medicine, University of Leeds, Leeds, UK; Thelethon Institute of Geneticsand Medicine, Naples, Italy Purpose Foveal hypoplasia associated with optic nerve misrouting has been exclusively associated with albinism. Here we describe FHONDA, a new syndrome consisting of Foveal Hypoplasia, Optic Nerve Decussating defect and Anterior segment abnormalities, in the absence of albinism. Patients and methods Fifteen affected family members of seven families, five of Asian and two of European origin, were clinically and genetically investigated. Clinical examination included fundoscopy, optical coherence tomography (OCT), and pattern onset visually evoked potentials (VEP) to determine misrouting. The gene mutated in FHONDA syndrome was mapped to a 3.1 Mb locus on 16q23.3–24.1. Sanger sequencing was used to screen all genes within the FHONDA locus. After identification of the mutated gene, further experiments were carried out to determine expression and to investigate the effect of the gene on ocular development and pigmentation. Results All patients had a poorly defined foveal avascular zone and absent macular and foveal reflexes. OCT showed a severe (grade 4) foveal hypoplasia. Pattern onset VEPs showed optic nerve misrouting. Eight patients from 3/7 families had anterior segment abnormalities, i.e. Axenfeld anomaly or posterior embryotoxon. None of the patients showed any pigmentation defects, such as iris transillumination, or reduced pigmentation

Doc Ophthalmol (2014) 129:13–55

27

in the skin, eyes and/or hair. Mutation analysis showed homozygous mutations in SLC38A8 in the five consanguineous Asian families. One European patient showed a hemizygous missense mutation combined with a large deletion; two sisters from the other European family had compound heterozygous mutations. SLC38A8 expression was found predominantly in neuronal tissue. Knockdown medaka fish embryos showed structural abnormalities of the eyes, but none of them were found to have any pigmentation defects of the retinal pigment epithelium or in melanophores. Conclusion utations in SLC38A8 cause foveal hypoplasia and optic nerve misrouting in the absence of pigment alterations. The precise role of SLC38A8 in retinal development and optic nerve decussation is not yet known. However, this study shows that the combined occurrence of optic nerve misrouting and foveal hypoplasia is not restricted to albinism. (12:15) The first reported two Egyptian families with posterior microphthalmos, retinitis pigmentosa and foveoscesis (MFRP syndrome) A. Shehab, S. Turk Ophthalmology Department, Minia University, Giza, Egypt Purpose To report the first two Egyptian families with positive consanguinity and MFRP syndrome. Method: Patients: two families are included in the study following diagnosis of the proband. For the first family, two siblings (a male and a female), their male cousin, and their parents were examined. The second family included one female and her parents. Comprehensive ophthalmic examination, complete electrophysiology testing, ultrasound (A & B scan) and OCT examination were performed in all affected family members. Polymerase chain reaction amplification of DNA obtained from peripheral blood lymphocytes and nucleotide sequencing of the complete MFRP gene were performed. Results The affected patients showed typical findings of the MFRP- related syndrome. A homozygote mutation in the MFRP gene was identified with genetic analysis in both families. Conclusion The two studied families with MFRP-related syndrome and their gene mutation are the first to be reported in Egypt.

PA6: Pediatrics Wednesday, July 23, 9:00–10:30 am

(9:00) The Electro-oculogram in Danon disease 1

2

1

1

D. A. Thompson , P. Constable , A Liasis , B. Walters , M. T. Esteban1,3 1

Clinical and Academic Department of Ophthalmology, Great Ormond Street Hospital for Children; 2Applied Vision Research Centre, City University; 3University College London Heart Hospital, London UK Purpose Danon disease is a rare X-L dominant disease caused by mutations in the lysosome-associated membrane protein-2

gene (LAMP2). It is characterised by cardiomyopathy, myopathy, and varying degrees of mental retardation. In the eye, LAMP2 is expressed only in the retinal pigment epithelium (RPE). This study investigates the previously-unreported impact of LAMP2 mutations on the electro-oculogram (EOG). Methods ISCEV standard EOGs to 30-degree saccades were recorded in a boy (age 13 years) with Danon disease, his mother (age 50 years), his maternal aunt (age 24 years) and female cousin (age 6 years). All have mutations in c294G [ A of the LAMP2 gene on Xq28. This leads to a stop codon at position 98 of the protein (p.Trp98X), by which none or aberrant protein will be formed. Other clinical ophthalmic examinations included full field ERGs, Goldmann kinetic visual fields and Spectral OCT with fundus autofluorescence (FAF) imaging. Results The EOG Arden Index of light rise:dark trough amplitude ratios appeared normal, ranging from 1.68 to 3.94, but these were spurious because the absolute dark trough voltages were abnormally low (median 140 lV, range 72–198 lV). Similarily the light rise amplitudes were reduced (median 297 lV, range 198–266 lV). Full field ERGs were just at the limits of normal, or showed very mild abnormalities in one patient (24 year old female). Fundus examinations revealed some diffuse, streaky pigmentary changes in the peripheral fundi of all patients, greater in the older females. Hyperfluorescent FAF changes were florid across the fundi of the two older patients, mild in the 6 year old female, but minimal in the 13 year old male who, in contrast, showed the most constriction of the central visual field. All but the youngest managed Goldmann visual field testing that showed a constriction of the central field to a red stimulus. The proband’s mother additionally showed an enlarged blind spot and N8 near acuity. The proband’s distance LogMAR VA with myopic correction was RE 0.12, LE 0.06, but near acuity was N8. His Ishihara colour vision was normal. Overall the monocular visual acuity ranged from 0.02 to 0.36 LogMAR. Conclusions The lack or loss of LAMP2 protein in Danon disease is associated with a decrease in the RPE trans- epithelium potential (TEP), which reduces the amplitudes of both the dark trough and the light rise recorded voltages of the EOG. The increased fundus autofluorescence may result from a breakdown of the natural maturation of lysosomes involved in both the autoand phagocytic pathways that clear shed outer segments and damaged organelles. The early aging changes highlighted by the FAF and the reduced EOG voltages indicate the RPE’s physiology is compromised and may be unable to maintain its structural integrity (tight junctions), resulting in reduced TEP and transport proteins, causing accumulation of lipofuscin.

(9:15) Screening and treatment for retinopathy of prematurity: an evaluation in the hunan province of China over an 8-year period P. Wang1, C. Bohua2 1

Hunan Children’s Hospital, Changsha, 2The Second Affiliated Hospital of Xiangya Medical School, Changsha, Hunan, China Purpose To describe the ROP screening and treatment data in Hunan province from 2006 to 2013 over an 8-year period.

123

28 Methods The purpose of this study was to analyze the annual ROP incidence, severe ROP incidence and blindness rate, and to compare the success rate and complications following cryotherapy surgery and laser ablation for ROP. Full-field ERGs were recorded following the International ISCEV Standard to evaluate the characteristics of the electroretinogram (ERG) in children with a history of pre-threshold or threshold ROP. A total of 23,475 premature infants underwent ROP screening. Of these, 13,289 cases were male, and 10,186 cases were female, with a gestational age (GA) from 26 to 37 weeks, with an average of 31.5 ± 3.8 weeks. The birth weight (BW) ranged from 500 to 3,000 g, with average of 1560 ± 850 g. For the first 4 years (from 2006 to 2009)screening was done using a binocular ophthalmoscope for 8,890 infants, and for the following 4 years (from 2010 to 2013), screening was done with a Retcam II for 14,585 infants. There was an average of 3.2 evaluations for each infant. Results We found 3,140 cases of ROP totally, with an incidence of 13.38 %: Four ROP infants had BW [ 2500 g, 88 ROP infants had BW [ 2,000 g, 390 ROP infants had BW [ 1,500 g. 7 ROP infants were diagnosed with birth age (BA) [ 34w; 34 infants were diagnosed for EO-ROP (early onset ROP). If the AAO criteria of ROP are used, the missing ROP rate was 390/3,140(12.4 %); using CMA criteria, 88/3,140 (2.8 %), but only 4/3,140 (0.12 %) according to current revised criteria of China. Surgery was undertaken in 536 infants, 221 patients were treated with cryotherapy surgery and other 315 patients with 532 nm green laser with extensive indications. The average time from birth to surgery was 11 weeks. Ultimately 459 cases were successfully treated after one surgical procedure, while 69 cases required multiple surgeries for skipped areas. There were 20 cases with retinal fold, and seven cases treated by vitrectomy, five cases resulting in retinal detachment. The incidence of ROP in males was higher than in females(v2 = 3.58, p [ 0.05); the incidence of ROP in BA C 32w infants was higher than BA \ 32 W infants (v2 = 6.18, p \ 0.05); ROP in BW C 1500G infants was more frequent than in BW \ 1,500 g infants (v2 = 6.70, p \ 0.01); the rate in single birth infants was higher than in multiple birth infants (v2 = 4.27, p \ 0.05). BW and GA of ROP infants were distinctly lower than non-ROP infants. The average BW of ROP neonates was 1,280 g (range 500–3000 g, SD 490.0 g). The average GA of ROP neonates was 29.0w (range 26–35w, SD 3.3w). A decreasing incidence of ROP has been observed during 2006–2010 and this has been attributed to the improving neonatal care and better monitoring of oxygen saturation in NICU in our province. There was a smaller increasing incidence since the introduction of Retcam II for screening in 2010, due to the advantages of Retcam II in diagnostic sensitivity and specificity, accuracy and easy for ROP screening. Surgery rate increased in 2011 due to revised indications for laser intervention. a-wave and b-wave of each response were observed. There was significant difference between ROP and control group in rod response (P \ 0.05), the latency was longer and the amplitude was lower in ROP group. But there is no difference between ROP and control group in cone response (P [ 0.05). Conclusions 1. Low BW(\1,500 g), low GA(\32w) and multiple births are high risk factors for ROP; 2. ROP screening criteria in different countries or regions should be modified and revised according to local multiple ROP center data. 3. Timely

123

Doc Ophthalmol (2014) 129:13–55 screening and surgery intervention decrease the blindness rate of ROP; 4. Compared to cryotherapy surgery, laser intervention offers lower risk and better prognosis; early intervention and repeated laser can improve the surgery outcome; 5. ERG in ROP children with history of pre-threshold or threshold ROP is abnormal compared to the children with history of prematuraty but without ROP, most evident in the rod response. (9:30) ERG characteristics in children with congenital nystagmus and early onset retinal dystrophies J. Brecelj1, A Kurent1,2, B. Stirn-Kranjc2 1

Unit of Visual Electrophysiology and 2Department of Paediatric Ophthalmology, Eye Hospital, University Medical Centre, Ljubljana, Slovenia Purpose To differentiate early onset retinal dystrophies on the basis of ERG characteristics in children with congenital nystagmus in comparison to age-matched controls. Methods Thirty-seven children with congenital nystagmus and early onset retinal dystrophies were included, with diagnosis according to clinical and ERG findings. Three ERG protocols were used according to the child’s age and 17 children were followed with two protocols: 27 (mean 2.1 years) were recorded with skin electrodes to flash stimulation, 16 (mean 6.5 years) with skin electrodes to full-field stimulation and 11 children (mean 12.2 years) with HK electrodes to full-field stimulation. ERGs were compared to age-matched controls, with differences considered significant if p \ 0.05. Results Clinical and electrophysiological findings were in agreement across all of the 37 studied children, 19 children were diagnosed by 2 years old, 10 children by 5 years old, and 8 children at between 8 years and 21 years. In nine children with Leber congenital amaurosis, the scotopic and photopic ERGs were not recordable under all protocols. Six children with congenital stationary night blindness had electronegative scotopic ERG under all protocols, and those with complete CSNB had absent rod ERG, and those with incomplete CSNB had reduced rod ERG. Eight children with achromatopsia had non-recordable photopic and subnormal scotopic ERG under all protocols. The implicit times of the scotopic b waves were prolonged. One child had blue-cone monochromatism, and reduced photopic and normal S-cone ERG. Six children with cone–rod dystrophy without systemic disorder and seven children with systemic disorder had affected photopic and scotopic ERGs under all protocols. Conclusions In children with congenital nystagmus, some early onset retinal dystrophies can be differentiated through the ERGs, even with skin electrodes.

(9:45) Full field ERGs in Individuals with Autism Spectrum Disorder P. A. Constable 1, S. B. Gaigg2, D. M. Bowler2, D. A. Thompson3 Applied Vision Research Centre, City University; 2Autism Research Group, City University; 3Clinical and Academic Department of Ophthalmology, Great Ormond Street Hospital for Children, Great Ormond Street, London, UK

Doc Ophthalmol (2014) 129:13–55 Purpose To investigate if the principle excitatory and inhibitory neurotransmitters implicated in the pathogenesis of autism spectrum disorders (ASD) cause changes to the scotopic and photopic ERGs in individuals with diagnosis of ASD. One earlier study found reduced scotopic b-waves in ASD and we wished to explore the ERG in greater detail in this population. Methods ISCEV standard ERGs were acquired to an extended range of flash strengths (0.001–200 cd s/m2) presented scotopically and 0.1–10 cd s/m2 photopically on a background of 30 cd/m2, from a group of ASD and typically developing (TD) subjects. The high functioning ASD group were diagnosed according to Diagnostic Statistical Manual-IV (Full IQ = 117 ± 11 and autism quotient = 35 ± 8). The 2 groups of 11 subjects were matched on age (p [ 0.6) [ASD = 36.9 ± 10.7 and TD = 34.5 ± 14.2]. All were free of ocular disease. There were no significant inter-ocular differences and the RE ERG data were used for comparison. Results There was no significant difference between groups in ERG time to peaks, but amplitudes showed some interesting trends; the ASD group showing overall smaller b-wave amplitudes whilst a-wave amplitudes are not significantly different between the groups. Results are reported as mean ± 2SD. For scotopic DA flashes, the rod-driven b wave amplitudes (lV) were DA 0.001 cd s/m2 ASD: 82 ± 69, TD: 109 ± 80, DA 0.01 cd s/m2 203 ± 118: TD 243 ± 146, DA 0.1 cd s m2: ASD 245 ± 142, TD 275 ± 141. For the scotopic rod driven DA 0.01 response, 3/11 ASD subjects (27 %) had response amplitudes more than 1SD below TD mean. DA 3 b-wave amplitudes from the ASD group were also smaller than those from the TD group [(DA 3.0 cd s/m2 ASD 310 ± 179, TD 361 ± 97; 10.0 cd s/m2 ASD 319 ± 224, TD 366 ± 91 and 200 cd s/m2 ASD 341 ± 245, TD 405 ± 110)]. The a/b wave ratios for the DA 1, 3, 10 and 200 cd s/m2 were not significantly different, although for two of the ASD individuals with low scotopic b-waves their a/b ratios were 10 % greater than the TD group average, suggesting that some individuals with ASD have abnormal ERG responses. The a-wave amplitudes (lV) for the groups were DA 10 cd.s./m2 ASD 220 ± 112, TD 242 ± 146 and DA 200 cd s/m2 ASD 289 ± 138; TD 336 ± 176. The photopic LA b-wave amplitudes in ASD subjects were also less than the TD group with the peak amplitude of the photopic hill less than the TD group [(LA 3.0 cd s/m2ASD 114 ± 62; TD 143 ± 78)]. Conclusions The results of this study find some b-wave amplitudes in the ASD group to be lower than the TD group. Several genes that contribute to neurodevelopment and neurotransmitters such as glutamate and gamma-aminobutryic acid have been implicated in the pathogenesis of ASD. Some of the differences in retinal responses in individuals with ASD may help identify abnormalities in specific neurotransmitters and their receptors that contribute to the ERG waveforms.

(10:00) A Simple method to obtain retinal responses from young children V. L .N. Fu1, A. Johnson1, K. K. Nischal1,2 1 Children’s Hospital of Pittsburgh, Pediatric Ophthalmology, 2University of Pittsburgh, Ophthalmology Department, Pittsburgh, Pennsylvania, USA

29 Purpose VEP and ERG provide an objective assessment of visual system. However, performing both tests on young children could be challenging. We want to find a method to screen retinal function without performing the standard ERG procedure. By this way, we may be able to detect more retinal diseases and uncover some underlying retinal disorders. Methods Eighteen children aged 1.1–16.3 years old (median age: 3.5 years old) underwent a standard ISCEV Flash VEP (FVEP) testing/protocol. At the same time, skin electrodes were placed under the lower eye lids to record retinal responses. Afterward, all children had full-field ERG compiled with the standard ISCEV ERG protocol. Seven children’ ERGs were recorded with corneal DTL electrodes and 11 children’s ERGs were recorded with skin electrodes. Results We found that waveform morphology was similar recorded with the two protocols. The average b-wave amplitude recorded using FVEP protocol was 2.5 ± 1.1 (skin electrodes) and 12.2 ± 5.6 (DTL electrodes) times smaller than averaged scotopic b-wave amplitude recorded using standard ISCEV ERG protocol. While the averaged a-wave amplitude from FVEP protocol was 3.2 ± 1.5 (skin electrodes) and 12.8 ± 3.8(DTL electrodes) smaller than the averaged scotopic a-wave amplitude from ISCEV ERG protocol. The b/a ratios were 1.9 (FVEP protocol with skin electrodes), 2.0 (skin electrodes with standard ERG protocol) and 1.9 (DTL electrodes with standard ERG protocol). Conclusions The ERG waveform morphology and b/a ratio are similar as recorded using the standard ISCEV protocol and skin electrodes. ERG responses are smaller when obtained with the Flash VEP protocol than using the standard ISCEV ERG protocol, either with skin or DTL electrodes. We suggest that recording the ERG with skin electrodes simultaneously with Flash VEP not only provides both cortical and retinal responses at the same time but also may provide a quick screening tool to detect retinal dysfunction in young children. However, establishing normative data for this method is essential for clinical diagnosis.

(10:15) Evaluating the effect of distance from the stimulator on the full field light adapted ERG OP responses J. Racine, R. Noles, Z. Peters Department of Ophthalmology, Nationwide Children’s Hospital, Columbus OH, USA Purpose In a clinical electrophysiology pediatric setting, it is often difficult to place a patient directly in front of the ganzfeld stimulator for proper recording of full field ERGs. In fact, a patient’s lack of cooperation sometimes requires the stimulator to be pushed further away in order to avoid potential harm. At other times, there is simply the inability of a patient to place his/her chin on a chin rest due to age or disability. Therefore, the purpose of this study was to investigate the effect of distance from the stimulator on the light adapted ERG OP responses, and to evaluate if increasing the flash intensity could compensate for the distance. Methods Light adapted luminance response functions were obtained from six (mean age: 26.83 ± 8.77 years old) subjects placed either directly in front of (0–5 cm), 10, 15

123

30 or 20 from the stimulator. Flash stimulations were provided by a ganzfeld stimulator (Color dome; Espion System E3) against a background light of 30 cd/m2. The flash intensities ranged from -2.9 to 3.0 log cd s/m2 with an inter-stimulus interval of 2 s. Results Independently of the flash intensity used, the a-wave amplitudes were always larger when the stimulator was closer to the subject. Overall, there was a 31.01 ± 12.73 % decrease in the a-wave amplitudes going from 0 to 5 to 10 cm, irrespective of the flash intensity. Similarly, there was a 24.60 ± 10.29 % decrease in the a-wave amplitudes from 10 to 15 cm and a 25.21 ± 6.38 % decrease from 15 to 20 cm at all intensities. In contrast to the a-waves, the b-wave amplitudes were not always larger when the stimulator was placed closer to the subject. In fact, at higher intensities ([1.69 log cd s/m2), the b-waves recorded at 0–5 cm were smaller compared to those recorded at 10 cm (p [ 0.05), 15 cm (p [ 0.05) and 20 cm (p [ 0.05). The light adapted b-wave Vmax was reached at 0.48 log cd s/m2 when subjects are placed directly in front of the stimulator, at 0.69 log cd s/m2 when placed at 10 cm (increase of 0.21 log unit; p \ 0.05) and at 2.3 log cd s/m2 when placed at 15 or 20 cm from the stimulator (increase of 1.82 log unit compare to 0 cm; p [ 0.05). Finally, the a- and b-wave peak times are not significantly affected by distance but there is a trend toward slower peak times with further distance. Conclusions The distance between the patient and the stimulator significantly affects the amplitude and to a lesser extent the peak time of the light adapted ERG and OP responses. The results obtained in this study also revealed that increasing the intensity of the flash stimulus does not compensate for the distance from the stimulator. Therefore, when working in a pediatric setting where the stimulator has to be placed further away, at times, from patients, it is strongly suggested that a normative database according to distance be established to better evaluate the electrophysiologic results obtained.

PA7: Methods 2 Wednesday, July 23, 11:00 am–12:30 pm

(11:00) Cone opponent and cone additive processing in incremental and decremental ERG responses J. Kremers, N. R. A. Parry, D. McKeefry, K. Focke, M. M. Jacob, T. Tsai Department of Ophthalmology, University Hospital Erlangen, Erlangen, Germany; Vision Science Centre, Manchester Royal Eye Hospital, Manchester, UK; Bradford School of Optometry & Vision Science, University of Bradford, Bradford, UK Purpose To describe L- and M-cone driven responses in incremental (On-) and decremental (Off-) electroretinograms. Methods L- and M-cone isolating stimuli were created using a four primary stimulator and the triple silent substitution technique (three photoreceptor types were not stimulated). Mean luminance and chromaticity were the same for L- and

123

Doc Ophthalmol (2014) 129:13–55 M-cone isolating conditions. The cones were stimulated using a sawtooth temporal profile: either ramp-down rapid-on for incremental (On-) responses or ramp-up rapid-off for decremental (Off-) responses. The measurements were performed at different temporal frequencies (2, 4 and 8 Hz) and with different stimulus sizes (full field and field diameters between 10 and 70 changed in 10 steps). Results The L-cone driven incremental and M-cone decremental responses resembled each other and were similar to the conventional On-responses (with homologues to the a- and b-waves). On the other hand, the responses to L-cone decrements and M-cone increments resembled each other and contained homologues of the d-wave. This was the case for all temporal frequencies. The antagonistic cone responses were particularly obvious at smaller stimulus sizes and were confirmed by implicit time measurements of the different response components. We also found that the implicit times, particularly of components homologous to the b-wave, increased with decreasing stimulus size. At full field, the responses to L-cone isolating stimuli were considerably larger than those to M-cone isolating stimuli. This response difference also decreased with decreasing stimulus size, suggesting that the far periphery (beyond 35 eccentricity) is strongly L-cone dominated. Conclusions The data strongly indicate that it is not sufficient to explain the origins of the ERG solely on the basis of On(depolarizing) and Off- (hyperpolarizing) cells. Cone signals are processed in parallel channels in a cone opponent or a cone additive manner through midget and diffuse bipolar cells. The activities of post-receptoral pathways are possibly also reflected in the ERG. Conventional ERGs to full field white light stimuli are probably mainly L-cone driven.

(11:15) Effect of pupil size on flicker ERGs recorded by new mydriasis-free ERG system (RETeval) M. Kondo, K. Kato, R. Nagashima, Y. Matsui, M. Sugimoto, and H Matsubara Department of Ophthalmology, Mie University Graduate School of Medicine, Japan Purpose To study the effect of pupil size on the amplitude and implicit time of 28.3 Hz flicker electroretinograms (ERGs) recorded by a new mydriasis-free full-field flicker ERG system (RETeval). This system was designed to deliver constant retinal illumination by adjusting the light intensity to compensate for the pupil size. Methods Ten healthy volunteers (age, 25–46 years) were studied. After 10 min of light adaptation, 28.3 Hz flicker ERGs were recorded every 3 min over a period of 21 min after starting pupil dilation by drops of 0.5 % tropicamide and 0.5 % phenylephrine. The pupil size was measured by an automated pupil area measuring device. The amplitudes and implicit times of 28.3 Hz flicker ERG were measured after the installation of the mydriatics. Results The average pupil area was 14.39 ± 4.78 mm2 at the baseline, and it gradually increased to 46.50 ± 12.20 mm2 at 21 min after starting the mydriatic drops. There was no statistically significant change in the amplitude of flicker ERG during the 21 min. However, the average implicit time of the

Doc Ophthalmol (2014) 129:13–55 flicker ERGs was 32.80 ± 1.03 ms at the baseline, and it significantly increased with time, reaching 34.20 ± 1.12 ms at 21 min (two-way ANOVA and Bonferroni type multiple comparisons). There was a significant positive correlation between the pupil area and implicit time (Pearson productmoment correlation: p \ 0.001, r = 0.62). A simple regression equation showed that a 1 mm2 increase of pupil area caused approximately 0.05 ms implicit time delay. Conclusions These results suggest that the implicit time of 28.3 Hz flicker ERG recorded with mydriasis-free flicker ERG System (RETeval) can be influenced by the pupil size.

(11:30) Comparisons of VEPs elicited by an organic LED (OLED) vs a cathode-ray tube (CRT) display V. Zemon1, J. G. Zweifach1, S. E. Sheldon1, K. B. Rothkopf1, M. Willoughby1, G. Hu2, C. Stewart3, J. Gordon4 1 Ferkauf Graduate School of Psychology, Yeshiva University, New York, USA; 2VeriSci Corp., Raritan, New Jersey; 3Konan Medical USA, City, California; 4Hunter College, City University of New York, New York, USA

Purpose Recent publications have addressed the properties of OLED displays in comparison to CRT displays and to liquid– crystal displays (LCD), to evaluate them for use in visual research. The conclusions drawn from this work are that OLED displays have excellent spatial, temporal, luminance, and chromatic characteristics, and therefore are suitable replacements for CRT displays which are no longer available in the marketplace. Specifically, it was noted that OLED displays should be used for eliciting pattern VEPs and ERGs, whereas LCD displays are not suitable for this purpose because of an inherent luminance artifact. In the current work, a battery of VEP tests was conducted with both a conventional CRT display and an OLED display to evaluate the veracity of this recommendation. Methods An EvokeDx system (Konan Medical) was used to present stimuli on either an NEC 17’’ CRT (VGA connection) or a Sony 17’’ OLED display (HDMI connection), and was used to synchronously collect and analyze the electrophysiological data. The luminance of both displays was set at *50 cd/m2 and the visual angle of the square screen subtended *10 deg. There were differences in the CRT and OLED displays in native spatial resolution (640 9 480 vs. 1920 9 1080 pixels) and temporal resolution (150 vs. 60 Hz frame rate). Look-up-tables were used to linearize the displays, and linearity was assessed by measuring luminance over time with a photodiode. Transient VEPs were elicited by a high contrast 32 9 32 checkerboard contrast-reversed with a 1 Hz square-wave signal in 2 s duration runs, and steady-state VEPs were elicited in swept-parameter paradigms with parametric manipulation of either spatial frequency (high-contrast horizontal gratings, 7.5 Hz contrast reversal) or contrast (16 9 16 isolated bright or dark checks appearing/disappearing sinusoidally from a static, uniform background at either 10 or 12.5 Hz). Each stimulus condition was repeated ten times to permit statistical analysis of the reliability of each type of response.

31 Results Fourier analysis of the photodiode signals demonstrated excellent linearity in the responses of both displays, with somewhat lower harmonic distortion and shorter response time with the OLED display. Transient VEPs were similar in waveform for the two displays, but latencies of the major peaks (N75, P100) were several milliseconds shorter with the OLED display, corresponding to its shorter response time. Magnitude-squared coherence values for distinct frequency mechanisms in the two sets of responses were also similar. These displays yielded similar contrast response functions (amplitude and phase of the fundamental frequency component) for bright or dark isolated-check patterns, with responses rising above noise level at about the same contrast levels, and spatial frequency functions with nearly identical grating acuity estimates. Conclusions The results confirm that the performance of the OLED display matches that of the CRT display in the generation of transient and steady-state VEPs to high and low contrast stimuli over a range of spatial and temporal conditions. Commercial relationships Hu and Zemon are principals in VeriSci Corp., which has a licensing agreement with Konan Medical USA; Gordon is a consultant and shareholder in VeriSci; Stewart is CEO of Konan Medical USA.

(11:45) Comparative analysis of the irradiance distribution in the rodent eye using either the Ganzfeld bowl or the Maxwellian view illumination technique N. A. Massie1, J. Rha2, N. Nestorovic2 Phoenix Research Laboratories, Pleasanton CA; Seattle Photonics, Seattle, WA, USA Purpose The objective is to determine the irradiance distribution in the rodent eye when using either the classical Ganzfeld bowl or Maxwellian view illumination approaches. Methods The question was addressed by both numerical simulations and analytical means. The numerical simulations utilized a widely respected optical design code ZEMAX and the analytical results examined the structure of the thick rodent lens. Results: Both approaches produced similar results. The illuminance distribution for the Ganzfeld bowl decreased with incident angle by cos(2q) whereas the Maxwellian view produced a uniform irradiance distribution. The impact of the steep decline in irradiance for the Ganzfeld bowl is increased when examined on an area weighted basis. The codes and analysis were also applied to the human eye, which features a thin lens, and the irradiance is essentially the same with either technique. Conclusions The use of the Ganzfeld bowl for animal eye studies should be questioned. The ERG signal can easily be a mix of both photopic and isotopic signal sources. And, when a calculation is made to determine photons per photoreceptor, this actual area illuminated is different than generally assumed. Commercial relationships Bert Massie and Jungtae Rha have a financial interest in the Maxwellian view Ganzfeld. Ned Nestorovic has no financial interest.

123

32 (12:00) Mobile device for long-term monitoring of CNS function via VEP testing M. Kuba, J. Kremlacek, F. Vit, Z. Kubova, J. Langrova, J. Szanyi, M. Chutna Department of Pathophysiology, Charles University in Prague - Faculty of Medicine in Hradec Kralove, Czech Republic Purpose All variants of the commonly used visual stimulators for VEP diagnostic applications cover significant portions of the visual field of the tested subject. The standard VEP examination also requires the subject’s visual fixation of the stimuli. Thus, the standard VEP examination does not allow any other parallel activity and it cannot be used for long term studies, e.g. for monitoring and evaluation of CNS functional changes or continual assessment of visual perception. Methods We tried to eliminate the above mentioned limitation by construction of a new visual stimulator (Czech patent 303192) consisting of 15 cm circular rows of 15 LEDs arranged on the margin of a peak (part of a baseball cap) that serves as a carrier for placement of stimuli in the periphery of the visual field. The assembly of LEDs enables generation of visual motion stimuli (sequential switching of single LEDs from the center to both sides or vice versa with random changes of directions) and flash stimuli (simultaneous switching of all LEDs with a selected frequency). Stimulation is controlled by a built-in microprocessor (motion stimuli have similar parameters as for clinical diagnostic purposes—see Kuba M et al., Vision Res 2007;47: 189–202). The new visual stimulator is combined with built-in EEG miniamplifiers and dry recording electrodes incorporated into the fixing belt of the peak (on the forehead at about Fp1 and Fp2 positions). Recorded VEPs are transmitted either via USB or ‘‘blue-tooth’’ to an evaluation notebook. Their significant changes can be also detected and automatically indicated to the tested subject with the use of an additional built-in processor. Results Motion-onset and flash VEPs can be recorded outside the visual cortical areas when appropriate stimulation is used in periphery of the visual field (outside central *20). Our arrangement enables the recording from the non-hairy (prefrontal) part of the head, which does not require any special montage of recording dry electrodes and VEP examination can start immediately after setting the peak on the head. Since the amplitude of VEPs to peripheral stimuli (dominant negative peak with latency between 200 and 300 ms) is even larger (even over 20 lV in some subjects) in prefrontal leads than in standard occipital locations, these CNS reactions seem to be rather event-related potentials than standard VEPs. Pilot tests with long lasting recordings of these VEPs (so far only in healthy subjects) have shown that their parameters are changing according to fatigue and emotional state. Conclusions Initial results with the use of our mobile device for VEP examination show that it could be useful not only for special diagnostic purposes (e.g. in immobilized patients, in intensive care units, for long-lasting CNS disorders) but it might also enable a long-term continual observation during activities in which a monitoring of brain functions is highly

123

Doc Ophthalmol (2014) 129:13–55 advisable, e.g. level of vigilance/fatigue in drivers and many other professions. Funding Charles University project PRVOUK—P37/07

(12:15) Full-Ffield chromatic pupillometry in the assessment of the Post-Illumination pupil response driven by melanopsin-containing retinal ganglion cells S. Lei, H. C. Goltz1,3, M. Chandrakumar1, A. M. F. Wong123 Program in Neuroscience and Mental Health, The Hospital for Sick Children; 2Department of Ophthalmology and Vision Sciences, The Hospital for Sick Children; 3 Department of Ophthalmology and Vision Sciences, University of Toronto, Toronto, Ontario, Canada Purpose A sustained pupil constriction can be observed after the offset of a bright blue light stimulus. This post-illumination pupil response (PIPR) is produced by the intrinsically photosensitive melanopsin-containing retinal ganglion cells (ipRGCs) and can be measured by chromatic pupillometry. While chromatic pupillometry holds promise as a new diagnostic and outcome measurement tool to assess ipRGC function, current testing protocols use central-field stimulation and require a very bright light of long duration which can be difficult for some subjects. We test the hypothesis that a more robust PIPR can be induced with full-field blue light stimuli of shorter duration and lower intensity than with existing protocols. Methods Ten visually normal adult subjects (mean age: 31 years, range 22–56) were tested. Pupil responses were recorded with an infrared eye tracker at 60 Hz. Full-field red (640 ± 10 nm) and blue stimuli (467 ± 17 nm) were presented in a darkened room using a Ganzfeld system. In experiment 1 (intensity trials), PIPR was induced using 1-second full-field stimuli of increasing intensity from 0.1 to 400 cd/m2 (11 steps). For comparison with a previously published protocol, a 60 9 90 central-field blue stimulus at 400 cd/m2 was also presented for 1 s. In experiment 2 (duration trials), PIPR was induced using 100 and 400 cd/m2 full-field stimuli of increasing duration from 4 to 1,000 ms (10 steps). Results PIPR increased monotonically with increasing stimulus intensity. Full-field stimulation using blue light at 400 cd/m 2 intensity induced significantly more sustained PIPR than central-field stimulation (p = 0.001, post hoc Student’s t test). In addition, PIPR increased as the duration of stimulus increased from 4 to 200 ms. However, no further increase in PIPR was observed when the duration increased from 400 to 1,000 ms. Conclusions Compared to existing central-field protocols, robust PIPR can be induced with a full-field protocol with lower intensity and shorter duration. This study is the first to demonstrate that saturating PIPR can be induced in vivo with a strong blue flash lasting only a few hundred milliseconds. This refined protocol will improve the recording quality and the subjective experience of future pupillometry testing.

Doc Ophthalmol (2014) 129:13–55

PA8: Toxicity and treatment Wednesday, July 23, 15:30–17:15

(15:30) Use of the mfERG in hydroxychloroquine retinal toxicity: defining earliest detectable onset and determining critical cumulative dose levels J. S. Lyons Georgetown University/Washington Hospital Center Program in Ophthalmology, Washington DC, USA Purpose To emphasize the importance of using the earliest reliable methods to detect hydroxychloroquine (HCQ) retinal toxicity, which can occur in asymptomatic patients with subtle or even unilateral changes in mfERG ring ratios and minimal visual field defects. Also, to further define the critical cumulative dose levels at which toxicity becomes an increasingly common event requiring more rigorous screening. Methods We have now evaluated over 800 patients referred for screening for Plaquenil toxicity. Screening included visual acuity, Humphrey visual fields, slit lamp examination, fundus examination, digital photography, multifocal electroretinography (mfERG), as well as full field electroretinography and SDOCT in some patients. mfERG testing was done using the Veris 61 Hexagon 4 min test strategy. To define the loss of paracentral function in HCQ patients characteristic of early toxicity, we compared the mfERG ratios of the P1-N1 amplitude of Ring 1 to that of Ring 2 and also of Ring 2 to that of Ring 5 with normal controls. Results Of the more than 800 HCQ patients screened in a tertiary clinical setting, we observed a variety of presentations of toxicity. The most common finding was, as expected, bilateral paracentral loss. But we also observed patients with paracentral loss combined with central loss, and patients with generalized loss as well as several patients with unilateral paracentral loss. In this group of 800 patients, we found 100 patients with HCQ retinal toxicity as defined by loss of mfERG paracentral amplitude by ring ratio analysis, visual field defects and clinical characteristics. All but 7 of these patients had taken more than 1,000 g of HCQ cumulative dose. Of those 7, three had taken Chloroquine in addition to Hydroxychloroquine, one had liver disease, and two received a very high daily dose. Our data show a prevalence of approximately 2 % of patients who have taken less than 1,000 g cumulative, 14 % of patients who have taken between 1,000 and 2,000 g cumulative dose, and over 27 % of patients who have taken over 2,000 g cumulative dose. Conclusion These data confirm our earlier findings that HCQ toxicity increases significantly with increasing cumulative dose beyond 1,000 g (7 years at 400 mg/day), but they also provide new evidence that beyond 2,000 g cumulative dose (14 years 400 mg/day) toxicity further doubles. Since many of these patients were referred for evaluation of newly discovered visual field defects, our prevalence data is likely to be skewed but not to the extent that it would obscure these very significant landmarks for distinguishing levels of toxicity. Our data also show that onset may occur even before symptoms are noticed and may first present in just one eye. At this stage we are most

33 likely to be able to prevent the development of symptoms and/ or to halt the progression of early retinal changes that have taken place. These findings should encourage the limiting of HCQ dose to minimum effective levels and to re-enforce the importance of vigilance in screening, particularly at high cumulative doses. (15:45) Sensitivity and specificity of SD-OCT outer retina changes in patients referred for Possible hydroxychloroquine toxicity R. Katira, J. Lyons Retina Group of Washington Georgetown University/ Washington National Eye Bethesda, MD USA Purpose To evaluate spectral domain -OCT (SD-OCT)findings in patients taking hydroxychloroquine. Methods 45 patients were referred for multifocal electroretinogram (ERG) testing and evaluation of possible hydroxychloroquine toxicity. Each patient was given a complete ocular exam, including vision testing, and dilated fundus exam. Patients also had high resolution SD-OCT, fundus photos, 10–2 Humphrey visual field and VERIS multifocal ERG evaluation using a Burian-Allen contact lens electrode. Results Twelve of the 45 patients exhibited SD-OCT abnormalities in the outer retina; particularly a gap in the inner and outer segment layers. Five patients were positive for hydroxychloroquine toxicity using multifocal ERG testing. Of these five patients, three exhibited early OCT changes (gap between inner and outer segment), one had more advanced foveal atrophy, and one had a normal OCT. Conclusions SD-OCT seems to be a useful tool in screening for hydroxychloroquine maculopathy. However, multifocal electroretinography is needed to definitively confirm hydroxychloroquine toxicity. (16:00) ERG monitoring of retinal function up to 5 years after completion of intra-arterial chemotherapy for retinoblastoma S. E. Brodie1,2, V. J. Mehta1, J. H. Francis1,3, B. P. Marr1,3, Y-P Gobin3,4, I. J. Dunkel5, D. H. Abramson1,3 1 Ophthalmic Oncology Service, Memorial Sloan-Kettering Cancer Center; 2Department of Ophthalmology, Icahn School of Medicine at Mount Sinai; 3Department of Ophthalmology, Weill Cornell Medical College; 4 Interventional Radiology, Departments of Neurosurgery, Neurology, and Radiology, New York-Presbyterian/Weill Cornell Medical Center; Ophthalmic Oncology Service, Memorial Sloan-Kettering Cancer Cente; 5Department of Pediatrics, Memorial Sloan-Kettering Cancer Center, New York, NY, USA

Purpose To provide extended follow-up of retinal function as monitored by ERG following successful completion of intraarterial chemotherapy for retinoblastoma. Methods This is a retrospective study of ERG outcomes based on 8 years’ experience following the introduction of intra-

123

34 arterial chemotherapy for retinoblastoma at Memorial SloanKettering Cancer Center. ERG recordings were obtained at examinations under anesthesia (EUA), generally at baseline (before initial intra-arterial treatment), 3 months after the final intra-arterial treatment, and annually thereafter. Most eyes also received focal laser or cryotherapy treatment subsequent to the final intra-arterial treatment. ERGs were recorded under sevoflurane anesthesia using ERG-jet contact lens electrodes and the DiagnosysLLC ColorBurst hand-held ganzfeld stimulator. Photopic responses were obtained in accordance with ISCEV standards. We report here the amplitudes of the responses to 30-Hz photopic flicker, which we have found correlate well with the entire sequence of responses in the ISCEV standard ERG protocol in these patients. Increments or decrements in flicker response amplitude less than 25 lV were regarded as insignificant. Responses at 1 year or more of follow-up were compared with responses at the 3-month EUA in order to allow for resolution of any acute effects of treatment. Results ERG data at baseline and 3 months after the final intraarterial treatment were available for 215 eyes in 190 patients. Of these, 55 eyes (25.6 %) lost 25 lV or more in flicker response amplitude at 3 months after treatment; 35 eyes (16.3 %) improved by at least that amount, with 125 eyes (58.1 %) not significantly changed. At 1 year after final intraarterial treatment, 14 of 131 eyes (10.7 %) lost significant ERG amplitude, 24 of 131 eyes improved (18.3 %), with 93 of 131 eyes (71.0 %) essentially unchanged. At 2 years, there were 12/69 (17.4 %) net losers, 16/69 (23.2 %) net gainers, and 41/69 (59.4 %) unchanged. At 3 years, there were 7/33 (21.2 %) net losers, 7/33 (21.2 %) net gainers, and, 19/33 (57.6 %) unchanged. At 4 years, there were 3/14 (21.4 %) net losers, 4/14 (28.6 %) net gainers, and 7/14 (50 %) unchanged. Of the three eyes for which we have 5-year data, 1 of 3 was a net gainer, 2 were unchanged. Conclusions ERG response amplitudes observed following completion of intra-arterial chemotherapy for retinoblastoma remained stable or improved in 80 % of patients during subsequent follow-up from one to 5 years. Approximately 1/5 of patients showed further improvement in ERG amplitudes after completion of intra-arterial treatment.

(16:15) How does ciliary neurotrophic factor affect the full-field ERG of normal human retina? B. L. Lam1, M. Liu1, W. J. Feuer1, W. Shi1, J. L. Goldberg2, A. M. Herro1, W. Tao1, R. Wen1 1

Bascom Palmer Eye Institute, University of Miami; Department of Ophthalmology, University of San Diego, USA

Doc Ophthalmol (2014) 129:13–55 photoreceptors and cells in the inner nuclear layer, this afforded a rare opportunity to investigate the long-term ERG effect of CNTF on the normal human retina. Methods ISCEV standard and classic full-field ERG responses were performed in 11 patients with POAG and 11 patients with AION who received CNTF-secreting implant in one eye. All subjects had normal appearing retina and normal full-field ERG responses in both eyes before unilateral CNTF implant placement. The ERG was repeated 1 year after CNTF implant placement. Results Compared to the fellow untreated eyes, the CNTF implanted eyes had significant decreases in scotopic rod ERG b-wave (mean change of -66 vs. +12 lV, p \ 0.001), in scotopic combined rod-cone b-wave (-77 vs -5 lV, p \ 0.001), and in total OP amplitude (-30 vs +10 lV, p \ 0.001). The cone flicker b-wave amplitude was decreased in the implanted eyes (-12 vs -5 lV, p = 0.11) although this is not statistically significant but the cone flicker b-wave implicit time was significantly prolonged (+1.2 vs -0.2 ms, p \ 0.001). The cone single-flash b-wave amplitude was not decreased in the treated eyes (-3 vs -8 lV, p = 0.39). However, the post cone b-wave waveforms of the treated eyes showed a notable significant morphologic increase in positivity above the pre-flash baseline amplitude, which was not present in the fellow untreated eyes. Analysis of the increased post cone b-wave positivity change of the CNTF treated eyes showed increases in i-wave amplitude (+18 vs +2 lV, p \ 0.001); in time-locked amplitude at 65 ms (+19 vs +5 lV, p = 0.065); in time-locked amplitude at 72 ms (+18 vs +2 lV, p = 0.004); and in maximal amplitude between 65 and 90 ms (+16 vs 0 lV, p = 0.004). Conclusions Our results showed CNTF-secreting implant reduces scotopic responses of the normal human retina, consistent with data from laboratory models. In addition, we found CNTF treatment induces a significant morphologic positivity change after the b-wave of the cone single-flash ERG response. This new observation indicates CNTF has a detectable ERG effect distal to the bipolar cells.

(16:30) Electrically elicited ERGs and VEPs in Argus II recipients: a window into signal processing in the degenerated RP retina G. Dagnelie1, H. C. Stronks2 1 Ophthalmology, Johns Hopkins University, Baltimore, MD USA; 2NICTA Canberra Research Laboratory, Canberra, Australia

2

Purpose Ciliary neurotrophic factor (CNTF) has been tested in several clinical trials (e.g. retinitis pigmentosa, macular telangiectasia. achromatopsia). In laboratory models, CNTF reduces rod and cone ERG responses. However, its effect on the normal human retina is unknown. We recorded ERGs from patients enrolled in a clinical trial aimed at the CNTF potential benefit to patients with primary open angle glaucoma (POAG) or non-arteritic anterior ischemic optic neuropathy (AION). Since the two diseases have no significant effect on

123

Purpose End-stage retinal degenerations such as RP show a rewiring of retinal tissue that fundamentally alters the normal localized, photoreceptor- and wavelength-specific signal processing streams that characterize normal vision. Retinal prosthesis systems such as the Argus II and Alpha IMS seek to convey visual information through localized stimulation of these cellular networks. We sought to characterize this altered visual processing through corneal recording of the electrically evoked electroretinogram (eERG) and scalp-recorded eVEP. Methods To explore the potential of the approach, we used Burian-Allen and occipital scalp electrodes to record the eERG

Doc Ophthalmol (2014) 129:13–55 and eVEP, respectively, in response to repeated pulse stimulation of all functional electrodes in 3 Argus II implant wearers. Subjects provided feedback by rating the brightness and size of the perceived flashes of light (‘phosphenes’). Input– output functions were generated using amplitude and latency based of the first two positive and negative peaks (P1, P2, N1, and N2). Correlations were determined using linear regression, followed by an F-test on the slope. Thresholds were defined as the amplitude equaling 4 times the standard deviation (4
SD) of the eVEP or eERG waveform. We also investigated the effect of stimulating different retinal locations, the maximal feasible pulse rate, and adaptation (‘fading’). Results eVEP waveforms had low signal-to-noise ratios, requiring long recording times and substantial signal processing. Waveforms varied between subjects, but showed good reproducibility and consistent parameter dependence within subjects. P2 amplitude was overall the most robust outcome measure and proved an accurate indicator of subjective threshold. Peak latencies showed small within-subject variability, yet their correlation with stimulus level and subjective rating were more variable than that of peak amplitudes. Pulse rates of up to 0.67 Hz resulted in reliable eVEP recordings. Perceived phosphene brightness declined over time, as reflected in P1 amplitude, but not in P2 amplitude or peak latencies. Stimulating-electrode location significantly affected P1 and P2 amplitude and latency, but not subjective percepts. For the eERG, waveforms were much more sluggish than for eVEPs, and no consistent correlates of peak amplitudes and latencies with stimulus amplitude were found. Conclusions eVEP responses show some variability between subjects, but do show consistent stimulus location and amplitude dependence. Corneally recorded eERGs, on the other hand, are much more difficult to characterize. We speculate that this is due to both the degree of disorganization in the degenerated retina and the unfavorable recording situation. Future retinal implants with recording capability through the stimulating electrodes will be needed to explore retinal activity following electrical stimulation at the local retinal level. Commercial relationship Second Sight Medical Products, Inc.-Center PI and consultant

(16:45) Visual outcome in 26 blind retinitis pigmentosa patients after receiving electronic subretinal implant Alpha-IMS E. Zrenner1, K. U. Bartz-Schmidt1, C. Chee2, F. Gekeler1, T. Jackson3, R. MacLaren4, J. Nemeth5, H. Sachs6, K. Sting1, D. Wong7 and The SUBRET Study Group 1

Institute for Ophthalmic Research, Centre for Ophthalmology, University of Tuebingen, Tuebingen, Germany; 2Department of Ophthalmology, National University Health System, Singapore, Singapore; 3Department of Ophthalmology, School of Medicine, King’s College, London, UK; 4Department of Clinical Neurosciences, Nuffield Laboratory of Ophthalmology, University of Oxford, Oxford, UK; 5Semmelweis University, Budapest, Hungary; 6Sta¨dtisches Klinikum Dresden-Friedrichstadt, Dresden, Germany; 7Department of Ophthalmology, University of Hong Kong, Hong Kong, Hong Kong. Purpose To assess the 12-month visual and safety outcomes following implantation of a 1,500-pixel subretinal implant in patients blind from retinitis pigmentosa (RP). This is the report

35 of a prospective mono-& multicenter clinical trial on safety & efficacy of subretinal implants for partial restoration of vision. (http://www.clinicaltrials.gov, Registration number: NCT00 515814, NCT01024803, NCT01497379) Methods Alpha-IMS subretinal implants (Retina Implant AG, Reutlingen, Germany) were positioned beneath the foveal region of 14 male and 12 female RP-patients (mean age 53.2 ± 8.2). Each of the 1,500 subfoveal photodiodes within an 11 by 11 field controls an amplifier that, depending on the strength of the light, emits currents to stimulate overlying bipolar cells. Power and control signals are supplied inductively via a subdermal, retroauricular coil from which a subdermal cable leads to the eyeball. Function was tested by 4 procedures: (1) Monitor-based standardized tests for light perception, light localization, movement detection, grating acuity and visual acuity with Landolt C-rings (2- or 4-alternative-force-choice-tests); (2) Detection, localization and identification of objects placed on a table; (3) Reading letters; (4) Visual experiences during outdoor and daily-life activities. Results (1) Implant-mediated light perception was possible for 22 (85 %) patients, light localization for 15 (58 %), movement detection (up to 35 cycles/degree) for 6 (23 %), measurable grating acuity (up to 3.3 cycles/degree) for 14 (54 %), and measurable visual acuity (up to 20/546) for 4 (18 %). (2) On a visual ability scale from 0 (worst) to 4 (best) for 4 white geometric figures presented on a black table, patients averaged 3.12 ± 0.31 for detection, 2.94 ± 0.3 for localization and 1.06 ± 0.28 for identification at month 2 which was significantly better (p \ 0.012) than with chip power switched off. Similar results were obtained with activities of daily living. Losses of approximately 1–1.5 U occured over 9–12 months. (3) Four patients (18 %) could read letters 4–8 cm in size. (4) Twelve patients (46 %) reported useful visual experiences including recognition of details, and 7 patients (27 %) could localize objects in daily life without details. (5) Besides two treatable serious adverse events there were no safety concerns. Conclusions The Alpha-MS implant has a CE mark for commercial use in Europe. Psychophysical testing and selfreported outcomes show restoration of useful vision in a majority of patients. Subretinal implantation surgery is safe and a multicenter study is continuing with a slightly modified implant for long time duration. Commercial relationship Study is supported by Retina Implant GmbH.

(17:00) Detecting early signs of hydroxychloroquine And Chloroquine retinopathy: A systematic review evaluating the multifocal electroretinogram as a screening rest under the 2011 revised American Academy of Ophthalmology recommendations S. Coupland1,2, S. Ahmadi1,2, A Tsang3, C. Gottlieb1,2 University of Ottawa Eye Insitute; 2Ottawa Hospital Research Institute, The Ottawa Hospital; 3Faculty of Medicine, University of Ottawa, Ottawa, Canada

1

Purpose To determine the sensitivity and specificity of multifocal electroretinography (mfERG) when used as a screening tool for detecting chloroquine (CQ) and hydroxychloroquine (HCQ) retinal toxicity. To evaluate the sensitivity

123

36 and specificity of mfERG in combination with other screening tests, such as automated visual fields (AVF), fundus autofluorescence (FAF) and spectral domain optical coherence tomography (sd-OCT). Methods A literature search was conducted for relevant publications from 2000 to 2013. Twenty-two qualified studies incorporating 422 eyes of 211 patients were analyzed. Using the criteria from the revised AAO recommendations, AVF, FAF, sd-OCT and mfERG data were compiled and classified as positive or negative test results. Dose [1,000 and \1,000 g was documented to stratify risk of retinal toxicity in patients on HCQ therapy. The sensitivity and specificity of mfERG was calculated following 2 models: using AVF as a comparative reference test, and a combination of two of three positive tests (AVF, FAF and sdOCT) as the comparative reference. The percent agreement between tests was calculated and an ANOVA was performed to determine if there was a difference in the rate of positive test results between the screening tests at a cumulative dose [1,000 and \1,000 g. Main outcome measures (1) Sensitivity and specificity of mfERG alone and in combination with other tests (AVF, FAF and sd-OCT) when used as a screening tool for detecting HCQ/ CQ retinal toxicity. (2) Sensitivity and specificity of mfERG, AVF, FAF and sd-OCT at doses of HCQ above and below the AAO-defined cumulative dose threshold for increased risk of retinal toxicity of 1,000 g. Results Using AVF as the reference test, the sensitivity and specificity of mfERG was 85 and 63 %, respectively. When agreement between at least two of AVF, sdOCT or FAF was used as the reference test, the sensitivity and specificity of mfERG increased to 97 and 91 %, respective;u. mfERG had the greatest agreement with sdOCT (94 %) and the combined reference test (93 %) followed by FAF (88 %) and AVF alone (75 %). mfERG had the greatest number of positive test results. The rate of positive test results differed significantly between the four screening modalities in patients exposed to a cumulative dose of less than 1,000 g of HCQ (F (3, 327) = 2.901, p = 0.035) and greater than 1,000 g of HCQ (F (3, 570) = 5.683, p = 0.0,008). Conclusions The mfERG had a good sensitivity and specificity in detecting HCQ/CQ retinal toxicity when used as a stand-alone test, however, the sensitivity and specificity were much better when mfERG was used with sdOCT or a combination of two of sdOCT, AVF or FAF. As a stand-alone test, FAF did not detect as many positives as mfERG, sd-OCT and AVF at doses above and below the threshold cumulative dose of 1,000 g of HCQ.

PA9: Optic Nerve and CNS Thursday, July 24, 9:00–10:30 am

(9:00) Electrophysiological characteristics in Leber hereditary optic neuropathy M. Hawlina1, M. Jarc-Vidmar1, J. Brecelj1, M Sˇusˇtar1, A Fakin1, B. S. Kranjc1, D. Glavacˇ2 1 Eye Hospital, University Medical Centre; 2Department of Molecular Genetics, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia

123

Doc Ophthalmol (2014) 129:13–55 Purpose To assess the electrophysiological characteristics of Leber hereditary optic neuropathy (LHON) in acute and later stages of the disease. Methods 8 patients with LHON were examined by means of Snellen visual acuity, Ishihara color vision, Goldmann or Octopus G2 TOP perimetry, fluorescein angiography (FAG), optical coherence tomography (OCT), electrophysiological and genetic testing, including whole mtDNA amplification and high-throughput sequencing. Results: Patients demonstrated typical LHON phenotype with bilateral poor visual acuity, abnormal color vision, central scotoma and hyperemic optic discs without leakage on FAG in the acute stage. Electrophysiology that was performed in 7 patients at the acute stage (average 5 weeks after onset of disease) showed abnormal N95 component of pattern electroretinogram (PERG) in 50 % and when followed (from 1 to 11 months) in 13/14 of eyes. In 4/7 eyes with abnormal PERG N95, temporal pallor of the optic disc was already visible whilst in the other 3 eyes, discs showed no pallor. All eyes with normal PERG N95 had hyperemic discs without notable temporal pallor. Electrophysiological assessment at follow-up (1–11 months) revealed an abnormal PERG N95 component in 13/14 eyes. The VEPs were abnormal in the acute stage in all of the affected eyes (in 12/14 eyes the P100 amplitudes were severely reduced and in 12/14 eyes P100 latencies were delayed) as well as at follow-up. In the only eye with normal PERG N95 at follow up, the optic disc was still hyperemic at 5 months after visual loss, while it was atrophic in all the others. OCT performed in the late stage of the disease showed severe optic atrophy in all patients (atrophy of retinal nerve fiber layer in 14/14 eyes). Conclusion PERG N95 loss and severely abnormal VEP may be present already in the acute stage of LHON, before disc pallor appears, suggesting primary dysfunction of ganglion cells.

(9:15) Electrophysiological characteristics of nonarteritic anterior ischemic optic neuropathy in a Chinese population S. Li, M. Wang, G. Wang, Y. Liu, Y. Wang, Z. Q. Yin Southwest Hospital/Southwest Eye Hospital, Third Military Medical University, Chongqing, China Purpose To investigate the electrophysiological and ophthalmic characteristics associated with nonarteritic anterior ischemic optic neuropathy (NAION) in a Chinese population. Methods Eighteen patients with a diagnosis of NAION underwent pattern reversal visual evoked potential (PVEP), flash visual evoked potential (FVEP) and pattern electroretinogram (PERG) examinations incorporating the ISCEV standards. The clinical findings were reviewed retrospectively and compared with the electrophysiological findings. Results Twelve patients were in acute phase (\1 month) while the other six patients were in chronic phase (\1 month). The majority of patients (10/12) of acute phase NAION showed decreased amplitude with normal implicit time for 1 PVEPs, and mildly delayed 0.25 PVEPs and FVEPs. PERG showed decreased P50 amplitudes with normal N95:P50 ratios. In the

Doc Ophthalmol (2014) 129:13–55 chronic phase NAION patients, 4/6 showed normal PVEPs, FVEPs and PERGs. Two of six showed delayed PVEPs and FVEPs, and loss of PERG N95, which were inconsistent with the optic nerve atrophy observed by funduscopy. Conclusions Patients with NAION have varied electrophysiological characteristics for PVEPs, FVEPs and PERGs with the different phases of the disease. The acute phase of NAION usually shows decreased amplitude with normal implicit time for PVEPs. The chronic phase of NAION may show recovery or atrophy of the optic nerve.

(9:30) The photopic negative response in Patients with Idiopathic Intracranial hypertension J. C. Park, H. E. Moss, J. J. McAnany Ophthalmology and Visual Sciences, University of Illinois at Chicago, Chicago, IL USA Purpose Idiopathic intracranial hypertension (IIH), defined by elevated intracranial pressure of unknown etiology, can produce irreversible vision loss by compromising optic nerve function. In the present study, the photopic negative response (PhNR) of the light-adapted full-field electroretinogram (ERG) was assessed in patients with IIH as a measure of inner-retina function. Methods Ten visually-normal control subjects (age 22–61 years) and 8 IIH patients with papilledema (age 19–52 years) participated in the study. The ages of the controls and patients did not differ significantly (t = 1.1, p = 0.30). Humphrey standard automated perimetry (SAP) 24–2 mean deviation (MD) scores were obtained from each patient. The PhNR was elicited by a brief, full-field, long wavelength (642 nm) pulse of light (0.2 cd.s/m2), presented against a short-wavelength (465 nm) rod-saturating adapting field (12.3 cd/m2). ERGs were obtained from one eye of each subject using standard recording techniques. PhNR amplitude was calculated as the difference between the baseline and the trough of the PhNR. For IIH patients who did not have a defined PhNR trough, response amplitude was calculated as the difference between the baseline and the amplitude at 77 ms, which was the mean implicit time of the PhNR of the control subjects. Results The mean PhNR amplitude of the patients (27.8 lV; standard error [SE] ± 6.6 lV) was significantly less (t = 2.3, p \ 0.05) than that of the controls (41.5 lV; SE ± 4.3 lV). The patients’ SAP MD values ranged from -4.0 to -25.6 dB and there was a significant linear correlation between MD and log PhNR amplitude (r = 0.84, p \ 0.05). Three patients whose PhNR amplitudes were within the normal range (from 39.6 to 53.8 lV) had only moderately abnormal MD values. The other five patients whose PhNR amplitudes were markedly abnormal (ranging from non-recordable to 17 lV) had relatively larger reductions in MD value. Conclusions The PhNR is an objective measure of retinal function that is highly correlated with perimetric visual field loss, assessed by SAP. The results suggest that PhNR measurement shows promise for assessing inner retinal function in IIH patients. Funding National Institute of Health grants R00EY019510 (JJM), K12EY021475 (HEM), and P30EY001792 (departmental core grant), Illinois Society for Prevention of Blindness, and

37 an unrestricted departmental grant from Research to Prevent Blindness

(9:45) Chromatic VEP gives additional information in young patients with demyelinating disease M. Tekavcic Pompe, B. S. Kranjc, J. Brecelj Eye Hospital, University Medical Centre, Ljubljana, Slovenia Purpose To compare chromatic VEP (cVEP) N wave and P100 VEP responses in young patients with demyelinating disease with or without optic neuritis (ON). Methods Thirty (30) patients (8–28 year) with demyelinating disease with or without ON were investigated. Electrophysiological evaluation consisted of standard pattern-reversal VEP and a new method, the cVEP. Two parameters were evaluated: cVEP N-wave latency and standard VEP P100 wave latency. cVEPs were recorded to an isoluminant redgreen stimulus. The stimulus was a 7 degree circle composed of horizontal sinusoidal gratings with a spatial frequency of 2 cycles/degree and 90 % chromatic contrast. Onset-offset mode of stimulation (On:Off = 300:700 ms) was used. The predominant part of the cVEP response in older children and adults consisted of a prominent negative (N) wave. Latency and amplitude of the N wave was analysed. Sixty (60) eyes were studied; 22 with at least one episode of ON (ON group) and 38 without any clinically evident episodes of ON (nONgroup). Results The average N wave latency in the ON group was 144 ± 44.5 ms, while in the nON group it was 117 ± 13.5 ms. The average P100 latency in the ON group was 116 ± 17.8, while in the nON group it was 102 ± 8.1 ms. There was no strong correlation between N and P100 wave latency (p = 0.57) in the ON group and also not in the nON group (p = 0.16). However, there were 4 eyes of 3 patients without clinical history of ON that showed markedly abnormal cVEP response (longer N wave latency and smaller N wave amplitude), but normal VEP P100. Conclusions The results of this study may indicate that cVEP can show more subtle changes in the parvocellular visual pathway function compared to standard VEPs. This is in concordance with studies showing that parvocellular visual pathway can be more often affected in demyelinating diseases.

(10:00) Occult optic neuropathy J. Heckenlively, K. W. Khan, K. R. Branham, C. Strong, K. Tozer, T. Jayasundera, T. Steffins Kellogg Eye Center, University of Michigan, Ann Arbor, Michigan USA Purpose To report diagnostic findings in patients with occult optic neuropathy, who commonly have no discernible optic nerve changes and have normal fundus examinations despite profound visual field loss. Methods Twenty-two patients who presented with strong complaints of vision loss had standard clinical and ISCEV ERGs performed. VEPs were performed in 11 patients. Retinal

123

38 and optic nerve head optical coherent tomography (OCT) was performed on all patients. Results All patients had markedly contracted Goldmann visual fields, normal ERGs, thickening of the peripapillary nerve fiber layer, and dysmorphic changes and thickened optic nerve heads (ONH). Many patients had a mass effect from the swollen nerve fiber layer on the ONH. The ONH changes observed on the OCT were not observable on the clinical examination. Without OCT imaging, a clinical diagnosis of optic neuropathy would been inconclusive. On review of the photopic ERGs, six patients had near normal (75–90 uV), 12 had normal (95–190 uV), and five had supernormal (195–300 uV) b-wave amplitudes. The visual field horizontal diameters were contracted from 15 to 30 with the IV-4e isopter. VEPs were normal to mildly abnormal. Discussion: With no obvious fundus or optic nerve findings, many patients are initially sent for MRI studies and when the MRI results are normal, they are referred to neuro-ophthalmologists or to psychiatrists for further examinations. The diagnosis of occult optic neuropathy is easy to identify with the combination of severely contracted fields, normal (or near normal) ERGs not consistent with the small fields, and an abnormal OCT scan showing dysmorphic optic nerve heads. The normal ERG and contracted visual fields suggest an optic neuropathy, but because standard clinical and electrophysiological testing has not been able to validate these patient complaints, patients end up seeing numerous doctors who doubt the validity of their complaints, with subsequent denial of insurance claims and emotional distress to the patient. (10:15) Evaluation of motion-onset VEPs in school-age children—comparison of normal readers and dyslexics Z. Kubova, M. Kuba, J. Kremlacek, J. Langrova, J. Szanyi, F. Vit, M. Chutna Department of Pathophysiology, Charles Univ. in Prague Faculty of Medicine in Hradec Kralove, Czech Republic Purpose Whilst in adults the motion-onset visual evoked potentials (M-VEPs) represent a useful diagnostic tool in neuro-ophthalmology (for review see Kuba M et al. Vision Res, 2007;47:189–202), it is rather difficult to evaluate this type of VEP in children (Kubova Z et al., Doc Ophthalmol 2014;128:121–129). M-VEPs mature over an extended time (up to 18 years of age), with gradual shortening of the main negative peak latency and VEP shape development (Langrova J et al. Vision Res 2006;46:536–544). Moreover, VEPs are unidentifiable to standard moving stimuli in some children. To describe this problem in more detail, we tested 19 dyslexics aged 7–12 years and 19 gender- and age-matched normal readers. Methods We performed the set of standard visual moving stimuli used in our lab for recording from adults—low contrast translation motion (TM) and expansion/contraction motion (EXCOM) (http://www.lfhk.cuni.cz/elf). Results In all tested children, reliable VEPs could be detected in 82 % to TM and in 95 % to EXCOM. In both M-VEP types, there was large inter-individual variability of the latency of the main negative (N2) peak (in normal readers 176–268 ms to TM and 176–266 ms to EXCOM; in dyslexics 206–326 ms to TM

123

Doc Ophthalmol (2014) 129:13–55 and 202 ms 326 ms to EXCOM). The group of dyslexics displayed significantly longer N2 peak latencies for both types of M-VEPs (p \ 0.01) than the group of normal readers. Pathological M-VEP latencies were found in 8 children to EXCOM and in 6 children to TM stimuli, in 3 children to both of them. Thus, electrophysiologic abnormalities were detected in 58 % of children from the dyslexic group. Conclusions Despite the fact that the large inter-individual variability of M-VEP latencies in children (based on variable maturation of motion perception) complicates the diagnostic use of M-VEPs in children, a distinct proportion of dyslexic children displays the reported ‘‘magnocellular deficit’’. Acknowledgement: Supported by the project of Charles University PRVOUK–P37/07.

PO1 Acquired retinal diseases 1-Unilateral cone-rod dysfunction associated with electronegative bright flash electroretinography T. Hirakata1, K. Fujinami124, Y. Kato1, 5, N. Nakamura1, T. Noda1,2, A Hirakata5, S. Ueno8, H. Ohguro9, Y. Miyake2,7, K. Tsunoda1,2 1 Department of Ophthalmology, National Hospital Organization, Tokyo Medical Center, Tokyo, Japan; 2 Laboratory of Visual Physiology, National Institute of Sensory Organs, National Tokyo Medical Center, Tokyo, Japan; 3Department of Ophthalmology, Keio University School of Medicine, Tokyo, Japan; 4UCL Institute of Ophthalmology, London, UK; 5Department of Ophthalmology, Tokyo Metropolitan Komagome Hospital, Tokyo, Japan; 6Department of Ophthalmology, Kyorin University School of Medicine, Tokyo, Japan; 7Aichi Medical University, Aichi, Japan; 8Department of Ophthalmology, Nagoya University Graduate School of Medicine, Aichi, Japan; 9Sapporo Medical University, Hokkaido, Japan

Purpose To describe two cases of unilateral cone-rod dysfunction associated with electronegative bright flash electroretinography (ERG). Methods A full medical history was obtained and a comprehensive ophthalmologic examination was performed, including visual acuity (VA), ophthalmoscopy, autofluorescence (AF) imaging, fluorescein angiography (FA), optical coherence tomography (OCT), visual field test by Humphrey field analyzer (HFA), electrophysiological assessment that incorporated the ISCEV standards full-field ERG(ffERG). Multifocal ERG (mfERG) was recorded in only case 2. Thirty-six causative genes for cone-rod dystrophy/retinitis pigmentosa were screened using the arrayed primer extension technology (APEX; Asper Biotech Ltd, Estonia). Immunoblot analysis, using the patient’s sera, was performed to detect autoantibodies against the retina. Results Case 1 was a 61-year-old Caucasian male who presented with VA loss since the age of 55 in the left eye (LE). He had complained of photophobia and night blindness in LE and had gradually noticed photophobia in the right eye (RE). The VA was 1.0 (RE) and 0.6 (LE). Ophthalmoscopy

Doc Ophthalmol (2014) 129:13–55 revealed multiple depigmented spots at the peripheral retina of each eye. No remarkable changes were demonstrated in AF, FA, and OCT in each eye. Decreased sensitivity of whole retina in LE and slightly reduced sensitivity in RE were detected by HFA. ffERG demonstrated severely affected responses in LE, with normal responses in RE. Dark adapted dim flash ERG (0.01 cd s/m2; DA 0.01) showed a remarkably reduced response and bright flash ERG (DA 30.0) had moderately reduced a-wave and severely reduced b-wave (electronegative configuration); severely affected responses were shown in light adapted cone ERG (LA 3.0) and 30 Hz flicker ERG (LA 3.0 30 Hz). No disease-causing variants were detected by APEX. Case 2 was a 49-year-old Japanese female with progressive VA loss in RE since the age of 48. She had breast cancer surgically resected at the age of 47. The VA at the initial examination was hand motion in RE and 0.02 in LE. Ophthalmoscopy revealed attenuated peripheral arteries in each eye, especially in RE, which was clearly detected on FA. No remarkable changes were demonstrated on AF or OCT. ffERGs were severely affected in RE and normal in LE. Dark adapted dim flash ERG showed a reduced response and bright flash ERG had moderately reduced a-wave and severely reduced b-wave (electronegative configuration). Cone ERG and 30 Hz flicker ERG were severely affected. mfERGs were grossly decreased in RE and normal in LE. FDG Positron emission tomography (PET) did not show abnormal uptake. Western blot analysis using the sera of each patient could not detect anti-retinal antibodies, including recoverin and TRPM1. Conclusions Two cases with unilateral cone-rod dysfunction associated with electronegative ERG were documented. Acquired retinal functional loss combined with negative ERG in cases with minimal fundus change suggests autoimmune retinopathy including cancer-associated retinopathy and melanoma-associated retinopathy. Specific retinal antibodies have not been revealed so far. Ischemic diseases could not be excluded entirely. Further systemic and immunological investigation is needed in order to resolve this disorder.

2- Electrophysiological examination of the patients following refractive surgery S. M. M. Shustarian Tehran medical branch, Islamic Azad University, Dr Shariati st Zargandeh, Beside Javahari Hospital, Tehran Medical Unit, Islamic Azad University, Tehran Iran Purpose Refractive surgery is among the ophthalmological procedures increasing in number, particularly in the younger generation. It is observed that some of the patients who have undergone refractive surgery complain of improper vision following the operation. On this basis, we planned to examine the visual system of patients following refractive examination using three electrophysiological tests: Visual Evoked Potential (VEP), Electroretinography (ERG), and Electrooculography (EOG) to evaluate the retina and visual pathway of these patients with normal ophthalmological examinations.

39 Method: Sixty patients following refractive surgery were selected randomly. The patients were examined using conventional electrophysiological tests 3 months after operation. The patients were classified into different groups according to sex, age and visual acuity before the operation. Conventional electrode attachment was used for this purpose. The Biomedica Mangoni instrument was used to record the VEP, ERG and EOG from the patients. The latency and voltage of the P100 VEP peak, parameters for b wave of the ERG, and the Arden Index of EOG was measured for all patients. SPSS software was used to analyze the results obtained. Results: It was observed that nine out of sixty patients had either subnormal VEP, ERG or EOG recordings. This result was observed mainly in older female subjects with low visual acuity before the operation. Conclusion From the result of present work one can conclude that it is necessary to examine the retina and visual pathway of older female subjects with lower visual acuity before refractive surgery using VEP, ERG and EOG tests.

3- Cataract surgery, a novel approach to readjust circadian clock in elderly? M. Wang1, X. Dong1, M. Liu2, W. Wang1 Department of Ophthalmology, Peking University Third Hospital, Beijing; 2 Alcon Laboratories, Inc. China Purpose Age-related ocular changes, for instance pupil miosis and opacity of the lens, reduce light transmission and circadian photoreception. By obtaining visual function and indicators of circadian rhythm in patients with age-related cataract before and after cataract surgery (phacoemulsification and achromatic intraocular lens implantation), the aim of this study is to investigate the influence of cataract induced blue light blocking in daytime alertness, diurnal melatonin secretion and sleep status in the elderly. Methods Twenty age-related cataract patients were recruited and their visual function, including best corrected visual acuity (BCVA), color vision, and contrast sensitivity, were estimated before and after cataract surgery. Electroencephalogram (EEG) measurements were collected continuously while subjects were exposed to blue (470 nm, 60lux, 20 lW/cm2) and red light (650 nm, 20lux, 13 lW/cm2) respectively 24 h prior to and subsequent to cataract surgery. Salivary melatonin levels were measured at 6, 7, 8, 12, 15, 21, 22, and 23 o’clock 48 h before and after cataract surgery by high performance liquid chromatography-mass spectrometry (HPLC–MS). Sleep quality was evaluated by Pittsburgh Sleep Quality Index (PSQI) 1 month before and 1 month after cataract surgery. Results The BCVA, contrast sensitivity, and color vision, especially in the blue-green to blue region of the spectrum, were significantly improved after cataract surgery (P \ 0.05). According to EEG records from 14 of the participants, exposure to blue light in the daytime resulted in decreased delta (0.25–3.00 Hz) and theta (3.25–7.00 Hz) activity, and increased alpha (7.25–13.00 Hz) and beta (13.25–30.00 Hz) activity compared to preceding dark conditions both before and after cataract surgery. The EEG showed no significant difference between anterior and posterior surgery in either blue or

123

40 red lighting conditions. Ten patients with undetectable levels of melatonin were excluded from melatonin data analysis. Salivary melatonin levels were increased, although not statistically significant, at 15, 21, 22 and 23 o’clock 48 h following cataract surgery compared to before surgery. The impact of cataract surgery on melatonin level was nearly statistically significant (P = 0.06). After the surgery, significant differences were detected in the melatonin level at 22 o’clock compared to the other six time points, except for 23 o’clock, while similar levels of melatonin were found among the various time points prior to the surgery. In addition, significantly improved daytime dysfunction was detected by PSQI 1 month after cataract surgery (P \ 0.05). Conclusion After cataract surgery, significantly improved visual function with increased blue light exposure of the retina was noticed in patients with age-related cataract. Diurnal exposure of blue light promoted alertness in age-related cataract patients and cataract surgery did not enhance daytime alertness based on EEG testing in the short-run. However, the surgery could possibly affect the regulation of systemic melatonin secretion and improve daytime function. Age-related cataract induced blue light blocking could play a role in sleep disorders in the elderly.

4- Macular structure affecting recovery of its function in patients with age-related macular degeneration after intravitreal ranibizumab T. Nishimura, S. Machida, K. Hashizume, D. Kurosaka Iwate Medical University School of Medicine, Japan Purpose To determine structural parameters that affect recovery of macular function in patients with exudative age-related macular degeneration (AMD) treated with intravitreal ranibizumab (IVR). Methods: We studied 30 eyes of 30 patients with exudative AMD who were treated with IVR 3 times at monthly intervals. The focal macular electroretinogram (fmERG) and spectraldomain optical coherence tomography (SD-OCT) recordings were made before and 3 months after starting IVRs. The fmERGs were elicited by a 15 white stimulus spot centered on the fovea. We measured each layer thickness of the fovea including the retinal layer, serous retinal detachment (SRD) and pigment epithelial detachment (PED). Parafoveal assessments were made on the horizontal and vertical meridians at 1.2 mm from the fovea, measuring thickness of the inner, middle and outer layers in addition to the SRD and PED. Correlation between these structural parameters and the a-wave amplitude of the fmERG was evaluated. Results There was no correlation between the structural parameters of the fovea and the a-wave amplitude. In the parafoveal region, the outer layer thickness significantly correlated with an increase of the a-wave amplitude (R = 0.562, P = 0.001). In addition, the SRD thickness was negatively and significantly associated with the a-wave amplitude (R = -0.539, P = 0.002). Change in the parafoveal SRD thickness by IVRs was the only independent determinant of recovery of the a-wave amplitude after treatment (P \ 0.05). Conclusions Macular function measured by the fmERG was determined by the parafoveal outer layer and SRD thickness in

123

Doc Ophthalmol (2014) 129:13–55 patients with exudative AMD. Out of these, changes of the SRD thickness by IVRs most strongly affected the recovery of the macular function.

5- Simultaneous recording of flash visual evoked potential and photopic electroretinogram in patients with vitreous hemorrhage Y-H Ohn, S. H. Lee, H. D. Kim Department of Ophthalmology, Soonchunhyang University Hospital, Bucheon, South Korea Purpose To investigate the usefulness of simultaneous recording of flash VEP and photopic ERG before treatment in patients with vitreous hemorrhage. Methods The ERG and VEP were recorded simultaneously in 13 eyes of 12 patients who had vitreous hemorrhage caused by proliferative diabetic retinopathy, retinal tear, and choroidal neovascularization. Intensity of light stimulus was gradually increased from 0.2 to 10 cd s/m2 in each eye to find stimulus thresholds of VEP and ERG. The thresholds of VEP and ERG responses were compared. All eyes underwent pars plana vitrectomy. Correlation between the stimulus threshold and postoperative visual acuity was analyzed. Additionally, correlation between the results of ERG and VEP at each stimulus intensity and the postoperative visual acuity was investigated. Results In two subjects, the stimulus threshold of the VEP was higher than that of the ERG. Postoperative visual acuity was worse than 20/200 in these patients. On the other hand, patients with a higher stimulus threshold of ERG than that of VEP demonstrated a better postoperative visual outcome. The results of ERG and VEP at 7.5 cd s/m2 stimulus intensity showed the highest correlation with postoperative visual acuity. Conclusion Severe media opacities including vitreous hemorrhage alter the intensity of the applied stimulus light and diminish the amount of light that reaches the retina. The results suggest that simultaneous recording of VEP and ERG may be a useful method to predict postoperative visual outcomes. A stimulus intensity of 7.5 cd s/m2 would be adequate in evaluating postoperative visual acuity in vitreous hemorrhage patients.

6- The changes of the multifocal electroretinogram after retinal detachment surgery G. Zhang1, H. Cheng2, R. Gao2, D. Wu2 1

Shenzhen Eye Hospital of Second Clinical Medical College of Jinan University, Shenzhen, China; 2 Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China Purpose To explore the changes of the multifocal electroretinogram (MERG) before and after retinal detachment surgery and to evaluate its clinical significance. Methods Twenty three patients with rhegmatogenous retinal detachment underwent MERG before and after surgery. The latencies and average response densities of a-wave and b-wave

Doc Ophthalmol (2014) 129:13–55 of MERGs in detached area, attached area, fovea, perifovea and the whole tested area were compared before and after surgery. Results Preoperatively, the average response densities of a-wave and b-wave in the detached area were significantly smaller than those in the attached area (t = 3.68, 4.26, p \ 0.01) and the latencies of a-wave and b-wave in the detached area were significantly longer than those in attached area (t = 3.07, 3.89, P \ 0.01). Postoperatively, the average response densities of a-wave and b-wave in the detached area, fovea, perifovea and whole tested area increased. There were significant differences between pre- and post-operation (P \ 0.05). However, the latencies of a-wave and b-wave before and after surgery showed no significant change (P [ 0.05). Conclusions mfERG is a useful tool in evaluating the recovery of posterior retinal function after retinal detachment surgery. The response density is more sensitive than the latency in testing retinal function.

Methods 7- A smart ERG signal generator for calibration of instruments and alignment of recording regimes across clinical laboratories M. Elm, A. C. Fisher, R. Laflin, R. Teymouri, A Eleuteri, R. P. Hagan Department Medical Physics & Clinical Engineering, Royal Liverpool University Hospital; Department of Physics, University of Liverpool, Liverpool, UK Purpose An ERG reference library using deterministic and stochastic models is available (AC Fisher et al., ISCEV Annual Symposium, Boston, 2014) that provides explicitly characterised recordings over a range of SNR’s with autoregressive continuous noise and spontaneous eye movement and blink artefacts. These digital records are freely accessible over the Internet from the Liverpool MatSOAP server using an MS Excel-based toolset for characterising ERG signal processing algorithms. In this presentation, the hardware and software of an ERG signal generator is described that can use these library records to drive the inputs of clinical ERG instruments. Methods A PC program written in MS Excel VBA and Matlab (Mathworks) was developed that downloads ERG reference library records with user-specified frequency content, SNR (ERG w.r.t. ARMA autoregressive noise), and number and amplitude of spontaneous noise artefacts arising from eye movements and blinks. A bespoke digital-to-analogue converter based on an ARM micro-controller is connected to the PC via a USB interface. The base bandwidth of the device is [0.1 … 1 k] Hz and the balanced output impedance 4.7kohm. Trigger synchronisation is provided by BNC hardwire connection and optical fibre. The output signals were available through standard touchproof 1.5 mm sockets compatible with current clinical instruments. Results ERG (clinically-based and synthetic) records were readily designed with explicit signal characteristics specified by the user using a modest MS Windows-based laptop acting as

41 a thin Internet client. ffERG, ssERG, PERG and mfERG recording scenarios were implemented and applied to a number of clinical instruments including the Roland Retiscan and the Diagnosys Espion. Conclusion The ERG signal generator provides test signals from an Internet-based reference library for the characterisation of electrodiagnostic instruments. It is suggested that such a device might be of use in calibration within quality control systems and alignment of recording regimes across clinical laboratories.

8- Characterization of exponential power spectral density decay of 30 Hz flicker in patients with retinitis pigmentosa R. Tzekov Department of Ophthalmology, University of South Florida, Tampa, FL USA. Purpose Light-adapted 3.0 flicker ERG (formerly ‘‘30 Hz flicker’’) response has been and is an integral part of the minimum ERG evaluation as part of the ISCEV standard and is widely used for assessment of photopic cone function in health and disease. Typically, this response is analyzed and reported in the time-domain by measuring amplitude (troughto-peak) and peak time (from stimulus to corresponding peak). Experimental studies have demonstrated that valuable information can be obtained also by spectral-domain analysis by examining the fundamental and the harmonics of the signal. Thus, the purpose of this work was to evaluate the power spectral density of 3.0 flicker ERG in patients with retinitis pigmentosa. Methods The light–adapted 3.0 flicker ERG records of 15 RP patients (8 M/7 F, age = 33.1 ± 20.8 years) were analyzed. Their records were compared to records of seven subjects with normal ERG (1 M/6 F, age = 33.7 ± 22.6 years). ERGs were recorded with UTAS E-3000 system (LKC Technologies Inc.) and DTL electrodes. Digitization rate was 2000 Hz and the first 396 samples of 512, representing the first 6 complete cycles of the recording, were selected for further analysis by Fast Fourier transform in Matlab (The MathWorks, Inc., Natick, MA). Fundamental power (at 30.3 Hz) and seven harmonic powers (from 60.6 Hz to 242.4 Hz) were analyzed. Results A simple exponential decay function model described well (R-squared [ 0.95) the change in power with increasing order of harmonics for a fundamental power of 2 microvolts or more, corresponding to an amplitude of *1.5 microvolts as determined by the LKC software (v.9.4.0) and a power of 4 microvolts or more at Y0 from the exponential model. The correlation between the fundamental power determined by FFT and the amplitude determined by the LKC software was 0.9186 for RP patients and 0.9043 for subjects with normal ERG, demonstrating good linear relationship with the fundamental power providing an estimate of amplitude which was *36 % lower for RP patients and *19 % lower for subjects with normal ERG. Various abnormalities in harmonic distribution were observed with amplitudes lower than 1.5 microvolts. Conclusions The results from this pilot study indicate that the fundamental and first seven harmonic powers of the light-

123

42 adapted 3.0 flicker ERG can be reliably obtained and analyzed by an exponential decay model for amplitudes greater than 1.5 microvolts. Further analysis will focus on informativeness of residual power in individual harmonics.

Doc Ophthalmol (2014) 129:13–55 Conclusions Pathological retinal conditions selectively affecting either rod or cone function enables us to visualize the contributions of each in the scotopic red response. References

9- Backyard electroretinography: a preliminary feasibility study S. E. Brodie Department of Ophthalmology, Icahn School of Medicine at Mount Sinai, NY, NY USA Purpose To explore the feasibility of recording ERGs using low-cost computer hardware and physiologic amplifiers. Methods ERG waveforms were simulated using the LKC Model 2506A Calibration System, which provides a 150 lV pulse of 30 ms duration in response to a flash of light. Recordings were attempted using the analog-to-digital convertor on a standard PC sound card, either with or without an additional front-end amplifier such as the Backyard Brains SpikerBox, or a researchgrade high-impedance, low-noise amplifier. Results Without front-end amplification, the PC sound card was just barely able to record the 150 lV pulse, with a signal to noise ratio of about 100 %. The SpikerBox was not suitable for ERG recordings, as it is too strongly differentiating to adequately display ERG waveforms. The waveforms obtained with the research-grade amplifier were very satisfactory. [The Raspberry Pi educational computer contains no analog-todigital converter, and so is unsuitable for this purpose without additional hardware.] Conclusions Direct ERG recordings are not feasible using only a PC sound card without additional front-end amplification. While the low-cost SpikerBox is not suitable for this purpose, it may be possible to develop an ERG recording system at low cost for demonstration purposes or for use in developing countries.

10- Scotopic red response: rod and cone components A. Chia, R. Png, L. Y. Chen, R. Mathur Singapore National Eye Center, Singapore Purpose To present the scotopic red responses in a variety of retina conditions. Methods In our lab, the scotopic red response (response to 2.5 cd.s/m2 red stimuli in the dark-adapted eye) is routinely performed between the scotopic and maximal steps of the ISCEV standard ERG. In this poster, a sample of responses ranging from normal to cone ± rod dystrophies (where cone function is selectively reduced), to a case of recovering vitamin A deficiency and rod-cone dystrophies (where rod function is reduced), and congenital stationary blindness (CSNB) are presented. This was to help us better understand how the response could be used clinically during ERG testing. Results In the normal response, there is an early x- and later b-wave response that correlates with cone and rod responses respectively. In eyes with reduced cone function, the x-wave is flattened. In eyes with reduced rod function and CSNB, the b-wave is flattened.

123

1. Creel DJ. The ERG and EOG: Clinical applications; http://webvision.med.utah.edu/book/electrophysiology/theelectroretinogram-clinical-applications/. 2. Lim SH, Ohn YH. Study of blue and red flash in darkadapted ERG. Korean J Ophthal 2005; 19: 106–11. 3. Verdon WA, Schneck ME, Haegerstrom-Portnoy G. A comparison of 3 techniques to estimate the human dark-adapted cone ERG. Vision Res 2003; 43: 2089–99.

11- The use of PERG in conjunction with mfERG in clinical cases R. Hagan, C. Jones, D. Jayakrishnan, A. C. Fisher, M. C. Brown Department of Medical Physics & Clinical Engineering, Royal Liverpool University Hospital, Liverpool, UK Purpose To review the use of PERG and mfERG in clinical cases, with particular emphasis on identifying when the macular tests gave different results. Methods The charts of all patients tested at the Royal Liverpool University between 1st Jan 2012 and 1st Jan 2014 were audited to determine how often the PERG and mfERG were used and how often they were abnormal. In particular, cases when one macular test was normal and the other was not were identified. PERGs were carried out to ISCEV standard. mfERGs were performed using a simplified procedure stimulating 19 segments in a 21 degree radius, scaled 4:1 and recording with gold foil electrodes for both. Results Of the 548 patients tested, about 70 % had one or both of the macular tests provided. The mfERG was used more often and tended to reveal an abnormality more often. The PERG was often used in conjunction with an abnormal PRVEP and whilst it may not have revealed an abnormality as often as the mfERG, it still added value to a number of cases by identifying that the macular optics were intact. Only 58 subjects had both PERG and mfERG, of those 12 had one normal test and one abnormal. Six had abnormal PERG with normal mfERG, these included patients referred with lattice dystrophy, left keratoplasty, query keratoconus, post cataract, and one who had a cataract removed post EDT with much improved vision. Six had abnormal mfERG in presence of normal PERG, these included a subtle macular problem and various field defects, including post diabetic laser and an enlarged blindspot. Conclusions The use of two macular tests is often useful in confirming or ruling out a diagnosis. Information from the two tests whilst largely generated from the macula has quite different mechanisms and so may not always agree. This audit of 2 years of ‘normal clinical’ practice would suggest that patients with media opacities may not always produce normal PERGs in the presence of normal mfERGs. Similarly, subtle macular abnormalities or off centre field defects may still have PERG within normal limits whilst showing an abnormality in the mfERG.

Doc Ophthalmol (2014) 129:13–55 12- Pattern visually evoked potentials elicited by organic electro-luminescence screen C. S. Matsumoto1, K. Shinoda1, H. Matsumoto2, H. Funada3, K. Sasaki1, H. Minoda1, T. Iwata4, A Mizota1 1

Department of Ophthalmology, Teikyo University School of Medicine, Tokyo, Japan; 2Matsumoto Eye Clinic, Tokushima, Japan; 3Engineering Department, Tomey Corporation, Japan; 4National Institute of Sensory Organs, National Tokyo Medical Center, Tokyo, Japan Purpose To determine whether organic electro-luminescence (OEL) screens can be used as visual stimulators to elicit pattern-reversal visual evoked potentials (p-VEPs). Method: Checkerboard patterns were generated on a conventional cathode ray tube (S710, Compaq Computer Co., USA) screen and on an OEL (17 inch, 320 9 230 mm, BVM-E170, Sony, Tokyo, Japan) screen. The time course of the luminance changes of each monitor was measured with a photodiode. The p-VEPs elicited by these two screens were recorded from 15 eyes of 9 healthy volunteers (22.0 ± 0.8 years). Results: The OEL screen had a constant time delay from the onset of the trigger signal to the start of the luminescence change. The delay during the reversal phase from black-to-white for the pattern was 1.0 ms on the cathode ray tube (CRT) screen and 0.5 ms on the OEL screen. No significant differences in the amplitudes of P100 and the implicit times of N75 and P100 were observed in the p-VEPs elicited by the CRT and the OEL screens. Conclusion The OEL screen can be used as a visual stimulator to elicit p-VEPs, however the time delay and the specific properties in the luminance change must be taken into account.

13- Organic light-emitting diode screens as visual stimulators to elicit multifocal ERGs K. Shinoda1, C. S. Matsumoto1,2, H. Matsumoto2, K. Seki1, E. Nagasaka3, T. Iwata4, J. Inoue5, A Mizota1 1 Department of Ophthalmology, Teikyo University School of Medicine, Tokyo,; 2Matsumoto Eye Clinic, Tokushima,; 3 Engineering Department, Mayo Corporation, Aichi,; 4 National Institute of Sensory Organs, National Tokyo Medical Center, Tokyo,;5Inoue Eye Hospital, Tokyo, Japan

Purpose To compare organic light-emitting diode (OLED) and cathode ray tube (CRT) screens as visual stimulators to elicit mfERGs. Method mfERGs were recorded from 7 eyes of 7 healthy volunteers (21 ± 2 years). The mfERGs elicited by a studio grade master OLED monitor (BVM-E170, Sony, Japan) were compared to those elicited by a conventional cathode ray tube (CRT) screen (S710,Compaq Computer Co., USA). The luminance changes of each monitor were measured with a photodiode. The CRT and OLED screens with frame frequency of 60 Hz were studied. A hexagonal stimulus array with 61 stimulus elements was created on each monitor. Results The luminance change showed that OLED had 9.2 ms of input lag compared to CRT stimulation. Since the duration of the on-luminance was short enough compared to the 16.67 ms in the OLED, no fusion with the consecutive onluminance was observed. The amplitude of P1 N1 of the first-

43 order kernels of the mfERGs were not significantly different between the CRT and OLED screens, while the implicit time was prolonged ca 10 ms mainly due to the input lag. Conclusion The OLED had good performance and we conclude that it can replace CRT as a stimulator for mfERGs; however normative data collection is recommended.

Glaucoma, Retinal Ganglion Cells, Optic Nerve, and CNS 14- Structure–function analysis in the follow-up of glaucoma: 2-flash mfERG (2F-mfERG), automated perimetry (AP) and Optical Coherence Tomography (OCT) A. M. Palmowski-Wolfe1, A. A. Ledolter1,2, M. Monhart3, A Scho¨tzau1, L. M. Brandao1 1

University of Basel, Department of Ophthalmology, Basel, Switzerland; 2Medical University of Vienna, Department of Ophthalmology, Vienna, Austria; 3Carl Zeiss Meditec, Feldbach, Switzerland Purpose To assess structure–function correlations in the follow up of primary open angle glaucoma (POAG) patients. Methods 10 eyes were included: 2 pre-perimetric (PPG), 2 high tension (HTG) and 6 normal tension (NTG) glaucoma patients. AP (automated perimetry): Octopus 101, G2; 2F- mfERG (VERIS 6.06TM, FMSIII) recording parameters: bandpass filter (BPF): 1–300 Hz, 103 Hexagons, M-sequence 213-1: LMax 100 cd/m2, Lmin \ 1 cd/m2, global flash: 200 cd/m2. In seven patients, an OCT was obtained: fast macular thickness protocol (Cirrus SD-OCT, Carl Zeiss). Analysis: AP: mean defect (MD in dB); 2F-mfERG: Responses filtered at 1–200 Hz. RMS was calculated for the focal flash response at 15–45 ms (DC) and for the global flash responses at 45–75 ms (IC1) and 75– 105 ms (IC2). Values were converted into logarithmic scale. OCT: full macular thickness (MT) and ganglion cell analysis (GCA = GCL + IPL) as per Cirrus Software. Statistical analysis was performed using the Wilcoxon test as well as linear mixed effects models in the statistical package R version 3.0.2. Results Mean group age was 64 years (SD ± 5.35) with an average time of 37.9 months (SD ± 8.0) follow-up. Mean IOP and BCVA were 11.3 mmHg (SD ± 1.76) and 0.99 (SD ± 0.03), respectively. During the follow-up period, mean MD changed from 4.39 dB (SD ± 3.76) to 5.15 dB (SD ± 4.66). From the first to the last visit, average MD increased by 0.76 dB (SD ± 1.61), while IC1 decreased by log 0.97 (SD ± 3.1) and IC2 decreased by log 0.63 (SD ± 2.8). Overall response of mfERG showed only a trend (p = 0.08), while response average in the central 10 degrees correlated significantly with changes in MD (p \ 0.001). Although the patient number was small, we analyzed the subgroups and saw that in PPG and NTG, but not in HTG, mfERG (overall \ central) correlated significantly to MD in IC1 and IC2. Significance levels for the central response average: IC1 (PPG: p \ 0.01 and NTG: p = 0.01) and IC2 (PPG: p = 0.02 and NTG: p \ 0.01). 7 patients from this sample had a followup period of 36.1 months (SD ± 8.5) with the Cirrus OCT. In these, mean MD increased by 1.0 dB (SD ± 1.6), IC1 decreased by log 0.29 (SD ± 3.5) and IC2 decreased by log

123

44 0.18 (SD ± 3.3) while MT increased by 1.27 lm (SD ± 6.6) but GCA decreased by 0.97 lm (SD ± 0.94). Conclusions Despite the small sample, initial results suggested a good correlation between MD and 2F-mfERG (especially IC1 and IC2) and especially in the central 10. While changes observed in the OCT are within the range of test–retest variability published in the literature, the slight decrease in the complex ganglion cell layer fits in with worsening MD and mfERG function in glaucoma. Acknowledgments AMP: SNF (Swiss National Science Foundation) NMS 1823, LHW Stiftung Lichtenstein.

15- The functional change of twelve patients with Leber’s hereditary optic neuropathy M Wang, G Wang, X Meng, Y Wang, Z Yin Southwest Hospital/Southwest Eye Hospital, Third Military Medical University, China Purpose To observe the characteristics of function damage in twelve cases with Leber’s hereditary optic neuropathy (LHON). Methods Twelve patients with the mean age of 28 years old (range from 13 to 59) were diagnosed as LHON by mitochondrial DNA testing. We compared the function test results of these subjects to normal values. Results Visual field testing (VF) in eight patients demonstrated visual field defects around the blind spot. Eight patients had Flash Visual Evoked potential (FVEP) testing, with implicit time of P2 waves all within the normal range. Seven patients had Pattern Visual Evoked Potential testing: 4 of them showed mildly prolonged P100 implicit time, and 3 had moderately prolonged P100 implicit time. Nine patients had Pattern Electroretinogram (PERG) testing and 6 of them showed reduced value of N95/P50 in both eyes while the P50 amplitudes were normal. Nine patients had Flash Electroretinogram(FERG) testing and the results were all normal. Eight patients had Multifocal Electroretinogram (mfERG) and six of them manifested decreased amplitude density within 10 of macula area but higher amplitude between 10 and 30. Two patients had reduced amplitude density of the entire 30 of the macular area. OCT in eight patients showed decreased ganglion cells and IPL complex layer in macular and retinal nerve fiber layer (RNFL) around optic disc. After the treatment with idebenone and coenzyme Q, two eyes of two patients showed mild improvement in BCVA, PVEP implicit time and mfERG amplitude density. Conclusions In these LHON patients, the decrease of N95/ P50 ratio in PERG showed damage of retinal ganglion cells, which was confirmed by the structural loss seen by OCT in the macular ganglion cell layer and in the RNFL around the optic nerve head. mfERG showed impaired retinal function in the macular area. The latency time of FVEP usually was not delayed, but the PVEP showed mildly prolonged P100 implicit times.

16- The left fusiform face area may play a role in dealing with information from the right one E. Wu, J. Chen, F. Lu School of Optometry and Ophthalmology and Eye Hospital, Wenzhou Medical University, Wenzhou, China

123

Doc Ophthalmol (2014) 129:13–55 Purpose study the different functions of bilateral fusiform face area (FFA) and the relationship between them. Methods Patients with problems with face perception were recruited primarily from patient populations of the neurology and neurosurgery department. Patients were tested with a facejudgment program, to determine the probability of their interpreting a non-face picture (NF0), a low-similarity face picture (NFL), a high-similarity face picture (NFH) or a FACE picture as a face. Using functional magnetic resonance imaging (fMRI), we assessed the functional effects upon and the relationships between the FFAs of presentation of pictures with different degrees of similarity to face pictures. Results A patient with a problem with face perception was identified, with a history of cavitary infarction, with the largest impaired area near the right FFA. For this patient, the probability of interpreting a low similarity face picture (NFL) as a face is about 0.1, compared with a probability of 0.05 for normal individuals. The probability of interpreting a highsimilarity face picture (NFH) as a face picture is about 0.28, greater than that of normal (0.11). The fMRI results show face stimuli lead to stronger activity in the right fusiform gyrus, but weaker activity in the left fusiform gyrus than stimuli depicting houses. Conclusions This patient’’s right FFA is apparently involved in face processing as a face-selectivity region, and his left FFA plays a role in the encoding and recognition of houses (instead of the parahippocampal place area (PPA)). According to this patient’s result, the left FFA is perhaps a region dealing with the signal coming from the right FFA and PPA.

Animal Models 17- A unique pattern of flicker ERGs distinguish coneand abnormal rod-driven responses in rd12 mice X. Dai1, J. Pang2 1

Eye Hospital, School of Ophthalmology and Optometry, Wenzhou Medical University, Wenzhou, Zhejiang, CHINA; 2Department of Ophthalmology, University of Florida, Gainesville, Florida, USA Purpose To explore a novel pattern of flicker ERGs that can distinguish cone- and abnormal rod-driven responses in rd12 mice: a spontaneous animal model of RPE65 Leber congenital amaurosis. Methods Flicker ERGs were recorded in wild-type C57BL/6 J (normal rod and cone function), Cpfl5 (cone photoreceptor function loss 5; pure rod function), Rho -/- (rhodopsin knockout; pure cone function), and Rpe65-deficient rd12 mice at postnatal day 16 (P16). Flicker stimulus light ranged in temporal frequencies from 5 to 35 Hz. Scotopic responses were elicited with dim (-2.0 log cd-s/m2) and bright (0 log cd-s/m2) light intensity. Photopic responses were elicited with +0.5 log cd-s/m2 light intensity. Double mutant rd12-Cpfl5 and rd12-Rho -/- mice were also studied to test photoreceptor origin of the remaining ERGs in rd12 mice. Results The ERG recordings to flicker yield a modulated response, which follows the stimulus. In the wild-type C57BL/ 6 J and Rho-/- mice (with normal cone function), bright flicker stimuli with both low-frequency (5 and 15 Hz) and highfrequency (25 and 35 Hz) could elicit a robust response, whereas

Doc Ophthalmol (2014) 129:13–55 the Cpfl5 (with normal rod function) and rd12 mice show only low-frequency (5 and 15 Hz) response. The ERGs of rd12 and double mutant rd12-Cpfl5 mice were almost identical, whereas there was no assessable response in rd12-Rho -/- mice. Conclusions Flicker ERG test is a simple and valid method of evaluating the respective contributions of retinal rod and cone systems. In Rpe65 deficient rd12 mice, the diminished high frequency flicker ERGs suggest the photopic single-flash ERG response is mainly from abnormal rods, but not cones.

18- Evidence suggesting that audition is relatively more impaired than vision following postnatal hyperoxia in rats X. Yang1,A. S. Lachapelle1,2, H. M. A. l. -Aamri4, Y. L. Ahn5, S. Chaychi1,2, S. Jung 2, S. Chemtob3,6, P. Lachapelle1,2 1

Department of Ophthalmology, Montreal Children’s Hospital, Montre´al, Quebec, Canada; 2Department of Ophthalmology, McGill University, Montreal, Que´bec, Canada; 3 School of Optometry, University of Montre´al, Montre´al, Que´bec, Canada; 4Sultan Qaboos University, College of Medicine and Health Sciences, Sultanate of Oman; 5Wonkwang University, Iksan, Republic of Korea; 6 Department of Pediatrics and Pharmacology, SainteJustine Hospital, Montre´al, Que´bec, Canada Purpose Severe visual and auditory impairments are often reported in preterm human infants. We compared the severity of both sequels using our rodent model of postnatal hyperoxia. Methods Newborn Sprague–Dawley rats (n = 17) were exposed to hyperoxia (80 % O2 for 22.5 h/day with three 0.5 h interruptions at 21 % O2) from birth (P0) to postnatal day 14 (P14), following which they were returned to normoxia (21 % O2). Control pups were maintained in normoxic condition throughout. Electroretinogram (ERG), Visual Evoked Potential (VEP), Auditory Brainstem Evoked Response (ABR) and Auditory Middle Response (AMR) were recorded at P22, P26, P30 and P60, respectively. Results At P22, the amplitude of the scotopic ERGs of hyperoxic rats were 60 % of normal and continued to decrease to reach 40 % of normal at age P60, compared to 40 % and 25 % for the photopic responses. Similarly, at P60 the amplitude of the VEP was 80 % of control. ABR tests revealed different degrees of injuries in exposed rats up to P30, the most severely affected showing less than 1 % of normal function as shown with significantly higher thresholds and markedly lower amplitudes of Peak II-V. The remaining rats had values similar to controls. However, by P60, all rats showed consistent hearing loss (about 5.6 % of control). Conclusions For the first time, this work demonstrates that postnatal exposure to hyperoxia in rats can lead to severe visual and hearing impairments. Our results would strongly suggest that audition is more affected in spite of a slight long-term recovery of function (compared to the gradual decline in ERG). Understanding the mechanisms underlying these phenomena can definitely help to avoid possible consequences in preterm infants.

45 19- An hypoglycemic Sulfonylurea shows novel retinal neuroprotective properties: pre clinical data M. Berdugo 1,2,3, M. Polak 1,4,5,6, E. KermorvantDuchemin1,2,3,4, K. Delaunay 1,2,3, M. C. Naud 1,2,3,, Z. Djerada 7, C. Gozalo 7, J-C Jeanny 1,2,3, C. Klein F. Behar-Cohen 1,2,3,8, P. Lassiaz 1,2,3

1,2,3

,

1

Universite´ Paris Descartes, Faculte´ de Me´decine Paris Descartes, Paris, France; 2INSERM UMRS 1138, Cordeliers Research Center, Paris France; 3Pierre et Marie Curie University (Paris 6), Paris France; 4Necker-Enfants Malades Children University Hospital, Pediatric endocrinology, Gynecology and Diabetology Department, Paris France; 5INSERM U1016, Paris, France 6Imagine Institute, Paris, France; 7Maison-Blanche Hospital University, Reims, France; 8Jules Gonin Ophthalmic Hospital, Lausanne, Switzerland Purpose To describe novel neuroprotective properties of a hypoglycemic Sulfonylurea, used in humans with Diabetes Mellitus, in rat models of retinal excitotoxicity and/or hyperglycemia-induced retinopathy. Methods Rat models of retinal neurodegeneration used are: intravitreal injection of N-Methyl D-Aspartate (NMDA)— neonatal hyperglycemia. Treatment consisted of orally administered or subconjunctivally or intravitreally injected Sulfonylurea (SU) versus vehicle, either before (prevention) or along with NMDA injection (treatment). Alterations and protection of retinal function were assessed by full-field ERG (in scotopic and photopic conditions, before vs. after retinal stress). Retinal cell death was quantified ex vivo by TUNEL assay and by histological cell quantification. An intraretinal cartography of the Sulfonylurea Receptor-1 (SUR-1) was established by immunohistochemistry. Ocular biodistribution of orally and intraveinously administered Sulfonylurea was assessed by High Pressure Liquid Chromatography. Results In the NMDA-induced retinal neurodegeneration model, alterations were partially avoided by Sulfonylurea treatment, with a dose–response profile. Basic (P \ 0.05 for prevention and treatment), scotopic b-waves (P \ 0.05 for treatment and P \ 0.01 for prevention) and scotopic OPs (P \ 0.05 for prevention and treatment). SU administered prior to NMDA-stress resulted in near normal ERG values (nonsignificant differences for basic, scotopic b-waves and scotopic OPs amplitudes). TUNEL assay showed very significant reduction of retinal cell death (P \ 0.001 for intravitreal preventive injection and sub-conjunctival treatment; P \ 0.01 for intravitreal treatment vs. control group). Ganglion cell layers of hyperglycemic neonates were histologically preserved (Control vs. hyperglycemic group: P \ 0.01; hyperglycemic vs. SU-treated hyperglycemic group: P \ 0.01; Control vs. SUtreated hyperglycemic group: NS). Ocular immunohistochemical cartography of SUR-1 revealed Sulfonylurea receptors in all internal retinal layers, from the inner to the outer limiting membrane. Its expression co-localized with glutamine synthetase labeling of the Mu¨ller glial cells. NMDA-induced retinal stress and SU treatment both modified expression and intraretinal localization of SUR-1, as well as retinal vasodilatation. Systemic SU administration

123

46 resulted in ocular biodistribution in the ocular media, which peaked at 1 h. Conclusion Pre clinical electroretinographic, immunohistochemical and histological results show that a Sulfonylurea treatment, administered by various routes, induces retinal neuroprotection. This suggests that it may be of interest in ocular pathologies involving ischemia and/or excitotoxicity such as glaucoma, diabetic retinopathy, and retinopathy of hyperglycemic neonates.

20- Cone-rich rodent Nile rats optimally model ON and OFF oscillatory potential properties in humans J. Gotzmann, I. S. Dimopoulos, Y. Sauve´ Department of Physiology, University of Alberta, Alberta, Canada Purpose To compare and contrast changes in the time and amplitude of the ON (b-wave) and OFF (d-wave) oscillatory potential (OP) components of the ERG. Methods Full-field ERG ON- and OFF- responses were recorded from 9 human subjects (aged 20–49 years) using DTL fiber electrodes and 10 Nile rats (Arvicanthis niloticus) using gold wire electrodes (under xylazine-ketamine anesthesia). Amplification (0.3–300 Hz bandpass) and digitization (1 kHz) were achieved with the Espion e2 system (Diagnosys LLC). Under 30 cd/m2 background adaptation, a white stimulus (intensity of 2.75 log cd/m2) was presented at durations increased in a stepwise fashion from 10 ms to 800 ms in eight steps. A total of 20 traces were averaged at each step. Individual oscillatory potentials were analyzed with trough-to-peak measurement from traces filtered at 75–300 Hz bandpass. Results Humans displayed four distinct OP peaks that were phase-locked to the ON response (ON-OPs). The ON-OPs (OP1-4) timing remained constant for all stimulus durations (peaks at t = 19 ms, t = 25 ms, t = 32 ms and t = 40 ms after stimulus onset) with amplitudes also remaining constant after stimuli durations [ 20 ms. For the OFF phase, two distinct OPs were distinguished. OFF-OPs (OP1-2) timings remained constant for stimuli durations [ 100 ms (peaks at t = 19 ms and t = 26 ms after stimulus offset). Nile rats displayed three distinct OP peaks that were phase-locked to the ON response. As with humans, timing remained constant for all stimulus durations (peaks at t = 20 ms, t = 27 ms and 39 ms). However, contrary to humans, the amplitude of the ON-OPs increased as a function of stimulus duration. For the OFF phase, two distinct OPs were also distinguished. The timing of the OFF-OPs was reduced as a function of stimulus duration. For both ON and OFF responses, OP peaks (from filtered traces) preceded the peaks from raw traces by 1–2 ms. Conclusions Our results demonstrate that Nile grass rats recapitulate human OPs more closely than traditional laboratory animals, such as mice and rats. As such, Nile rats represent a promising choice for studying human retinopathies.

123

Doc Ophthalmol (2014) 129:13–55 21- Females age better than males, especially when stressed! S. Chaychi1, X. Yang1, S. Chemtob2, P. Lachapelle1 1 Department of Ophthalmology & NeurologyNeurosurgery, McGill University - Montre´al Children’s Hospital Research Institute,; 2Pharmacology, McGill Univ/ Montreal Children’s Hosp, Montreal, QC, Canada

Purpose Aging is the single most important source of cumulative oxidative stress. The reactive oxygen species thus produced could potentially damage organs and tissues, including the retina. It would appear that females are more resistant to the aging processes, presumably due to the antioxidant effect of estrogen. The purpose of this study was to investigate the gender differences in retinal structure and function during the normal aging process as well as examine the effect of estrogen on the retinal function and structure in response to a lightinduced oxidative stress. Methods The retinal function of 50 normal male and female Sprague–Dawley (SD) rats was investigated with the ERG at postnatal day (P) 30, 60, 100, 200, and 300 (n = 5–7 male and female rats/age). In parallel, light-exposed (Light-exposed: LE; 10,000 lux; from P14-P28) male SD pups were injected with 17 a-Estradiol (E2) (4 and 12 lg, IP, daily from P12-P28). Experimental rats were subdivided as follows: Group 1 (G1: LE + 4 lg E2; N = 8); Group 2 (G2: LE + 12 lg E2; N = 8); Group 3 (G3: control + 4 lg E2; N = 7); Group 4 (G4: control + 12 lg E2; N = 7); Group 5 (G5: LE only; N = 8); Group 6 (G6: control; N = 7). ERG and histology were obtained at P30. Results ERG analysis showed that, irrespective of age, the average amplitudes of the scotopic a-wave and scotopic b-wave (normalized to P30) of female rats were higher than the males [71.45 vs. 60.81 % (P \ 0.032); 76.44 vs. 70.73 % (p = 0.2), respectively], while males had higher amplitude of the photopic b-wave compared to females (77.72 vs. 70.27 %). The scotopic and photopic ERGs of exposed rats were significantly lower than control groups (P \ 0.05). Among exposed animals, G2 showed higher amplitudes for all ERG parameters compared to G5 (scotopic a-wave: 73.7 ± 27.2 lvolts vs. 58.5 ± 19.2 lvolts; scotopic b-wave: 392.9 ± 129.4 lvolts vs. 344.9 ± 64.2 lvolts; photopic b-wave: 101.7 ± 29.9 vs. 85 ± 23.3 lvolts). Interestingly, the ERG amplitudes in G2 were higher compared to G1 (scotopic a-wave: 73.7 ± 27.2 vs. 59.1 ± 24.9 lvolts; scotopic b-wave: 393 ± 129.4 vs. 309.8 ± 84.8 lvolts; photopic b-wave: 101.7 ± 29.9 vs. 78.1 ± 18.8 lvolts), suggesting a dosedependent effect. Retinal histology disclosed a significantly thinner retina in LE rats compared to controls (P \ 0.05). Interestingly, the outer nuclear layer thickness (the most affected layer in light-induced damage) was remarkably preserved in exposed group treated by 12 lg 17 a-estradiol compared to the untreated LE rats [13.2 ± 3.2 vs. 5.7 ± 1.2 lm (P = 0.0009)]. Conclusions Our results demonstrate that female retina has a better tolerance to oxidative damage, whether the latter results from the normal aging process or the consequence of a pathological exposure to an exogenous oxidative stress. Our results would suggest that estrogen, with its

Doc Ophthalmol (2014) 129:13–55 demonstrated antioxidant effect, would explain this sexual difference. Funding CIHR.

47 2- Electrophysiological, genetic and ophthalmological findings in children with Leigh Syndrome

PO 2 Pediatrics

˚ kebrand1, K. Sofou2, S. Andersson1, R A A. K. Seyedi Honarvar1, N. Darin2, M. Tulinius2, M. Andersson Gro¨nlund1

1- Melanoma and Cancer Associated Retinopathy (MAR and CAR syndrom): two physiopathological and electrophysiological distinct entities to be aware of—report of two cases

1 Inst of Neuroscience and Physiology/Ophthalmology, The Sahlgrenska Academy at the University of Gothenburg,; 2 Dept of Pediatrics, Institute of Clinical Sciences, The Sahlgrenska Academy at the University of Gothenburg, Sweden

L. Abouaf0, R. Sallit1,5, L. Adelaı¨de2,5, A. M. Nguyen3, P. Se`ve2,5, J. Honnorat4, D. Biotti15, A Vighetto1 0

Neuro-Ophthalmology Department, Hoˆpital Neurologique Pierre Wertheimer, Hospices Civils de Lyon, Bron, France; 1 Department of Neuro-ophthalmology, Hoˆpital neurologique Pierre Wertheimer, Bron, France; 2 Department of Internal medicine, Hoˆpital de la Croix Rousse Lyon, France; 3Department of Ophthalmology, Hoˆpital de la Croix Rousse Lyon, France; 4Department of Neuro-oncology, Hoˆpital neurologique, Bron,; Center for Research in Neuroscience(CRNL), team of neuro-oncology and neuro-inflammation (ONCOFLAM) INSERM, Universite´ Lyon 1, France; 5Universite´ Claude Bernard Lyon 1, France Purpose Paraneoplastic retinopathies are a well-known diagnosis though it is rare. However the diagnosis can be difficult or delayed. We aim to describe two cases (one MAR and one CAR) which had very distinct ERG recordings. Methods We will show full ophthalmological evaluation in relation with ERG recordings (including visual fields and fundus photographs). Results The patient with MAR syndrome had a typical electronegative ERG with no ON response at the ON–OFF ERG recording while the patient with CAR syndrome had diminished amplitudes with mostly prolonged latencies. The first patient (MAR syndrome) had a history of melanoma treated 3 years ago and complained of visual field constriction and photopsias and was improved after intravenous immunoglobulins. No recurrence of the melanoma was found more than 1 year after the beginning of symptoms. Unfortunately, no antibodies were found after testing the serum on a retina of rat. The second patient developed bilateral rapidly progressive tunnel vision leading to the diagnosis of CAR Syndrome. The abnormal ERG recording, associated with asthenia and cervical adenopathy, led us to search for the cancer. An Undifferentiated Carcinoma of Nasopharyngeal Type (UCNT) was revealed thanks to PET (Positron Emission Tomography) scanner and biopsies. Anti-retinal antibodies were detected in blood serum. Conclusions Recognition and diagnosis of paraneoplastic retinopathies can be of major help in the search for cancer and the initiation of treatment as early as possible. ERG is an indispensable test in the diagnosis of abnormalities of the visual field with normal fundus and can be of various patterns depending on the aetiology (MAR or CAR). This is the first report of CAR associated with undifferentiated carcinoma of nasopharyngeal type.

Purpose Leigh syndrome is an inherited progressive mitochondrial disorder that affects the central nervous system with an onset in early childhood. The major clinical features are failure to thrive, hypotonia, psychomotor regression and brain stem dysfunction. The condition can be caused by several different gene mutations in mitochondrial or nuclear DNA. The aim of this study was to describe ophthalmological outcome in relation to electrophysiological and genetic findings in children with Leigh Syndrome. Methods A retrospective study was performed on 43 children with Leigh syndrome (19 females) with a median age of 2.4 years (range 0.7–14.2), investigated and diagnosed by a multidisciplinary team at the Queen Silvia Children’s Hospital in Gothenburg, Sweden, during 1987–2012. Nineteen were diagnosed with a genetic mutation. All the children received an ophthalmological examination, including visual acuity (VA), eye motility, refraction (expressed in spherical equivalent and adjusted for age), ophthalmoscopy and, in 22 of the cases, a fullfield ERG. In 17/22 children, ERG was performed under anesthesia, during the same session as a muscle biopsy was taken. Five children performed ERG awake, on three of whom skin electrodes were used. ERG was performed in one of the eyes. Results 30/43 (70 %) of the children had one or more ophthalmological finding such as low VA (n = 9/41), myopia (n = 2/23), hyperopia (n = 6/23), astigmatism (n = 12/23), ptosis (n = 3/19), reduced eye motility (n = 5/24), strabismus (n = 6/24), nystagmus (n = 4/24), photophobia (1/4), optic atrophy (OA) (n = 2/20), or pigmentation in macula/ periphery (n = 4/18). 17/22 investigated children with fullfield ERG were found to be normal. Two children had cone dystrophy, two children had rod-cone dystrophy and another child had reduced rod amplitudes. Of the children with normal ERG, 5/16 had a low VA, 2/17 had OA and 3/17 had pigmentation in the macula, while the children with affected ERG had a low VA (3/3 children), OA in 2/5 and macular pigmentation in 1/4. Eight genes were involved in the 19 children with known gene-mutations. No relation was found between the genes involved and ERG/ophthalmological findings. Conclusions In this unique cohort of children with Leigh syndrome, a vast majority showed ophthalmological findings and pathological ERG was found in 5/22 children investigated. The ERG findings showed a variety of abnormalities with no relation to clinical phenotypes or genotypes. Thus, we recommend an ophthalmological examination including ERG in all children with Leigh syndrome.

123

48 3- Electroretinogram changes of retinopathy of prematurity after photocoagulation H. Cheng, G. Zhang, F. Wen, S. Huang, D. Wu Shenzhen Eye Hospital of Second Clinical Medical College of Jinan University, Shenzhen, China; Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China Purpose The effects of photocoagulation on visual function in the eyes with retinopathy of prematurity (ROP) were studied. Methods ERGs were recorded with Espion electric physiological instrument (Produced in US) in 24 ROP cases (48 eyes) after photocoagulation and 20 non-ROP cases (40 eyes). B-wave amplitude and latency of rod response, a-, b-wave amplitude and latency of mixed reaction and cone responses were compared between the two groups. Results B-wave amplitude of the rod response was 95.76 ± 45.18 lv in the photocoagulation group and was 111.70 ± 48.12 lv in the control group. There was a significant difference between the two groups (t = 2.52, p = 0.029). B-wave latency was 107.36 ± 8.27 ms in the photocoagulation group and was 100.81 ± 11.79 ms in the control group. There was a significant difference between the two groups (t = 2.69, p = 0.021). The amplitude and latency of a-, b-wave in the mixed reaction and cone responses has no significant difference between the two groups (p [ 0.05). Conclusions Photocoagulation for retinopathy of prematurity results in damage to rod cell function, but had little effect on cone function 4- Visual evoked potential study in premature and term infants H. S. Chin1, S. J. Han2 Ophthalmology, Inha University, Incheon; 2Department of Rehabilitation Medicine, Ewha Womans University School of Medicine, City?, South Korea Purpose The visual evoked potential (VEP) in infants is used as a noninvasive method useful in diagnosing the disease of the optic chiasm and optic nerve. As preterm infants are especially vulnerable to brain injury induced by hypoxia of ischemia, VEP study is used often as screening test. The purpose of this study is to evaluate the relationship between the visual pathway abnormality and VEP study results in term and preterm infants. Methods Flash VEP studies of 58 term and preterm infants requested at the rehabilitation medicine center were analyzed retrospectively. On the basis of the findings of brain CT and MR images, the infants were divided into two groups, one with a lesion in visual pathway (retina, optic nerve, optic chiasm, lateral geniculate body, optic radiation, and visual cortex) and the other with no such lesion. The results of VEP studies were scored ‘‘0 score’’ in case of showing normal P2 latency, normal waveform and good reproducibility, and ‘‘1 score’’ in case of showing attenuated waveform or delayed P2 latency, and ‘‘2 score’’ in case of no response. Statistical analysis was based on the Mann–Whitney U test. Results 42 term infants younger than 5 were studied (boys: 32, girls: 10, mean age: 61.27 ± 516.4 weeks, age range: 2 days– 5 years), and 16 preterm infants younger than 38 weeks (boys:

123

Doc Ophthalmol (2014) 129:13–55 7, girls: 9, mean gestational age: 41.75 ± 9.7 weeks, range of gestational age at study: 33–66 weeks, range of gestational age at birth: 25–38 weeks). Twelve term infants out of 42 and 5 preterm infants out of 16 had lesion in the visual pathway. In preterm infants, there was no significant difference in VEP scores between preterm infants with visual pathway abnormality and without abnormality (right p = 0.52, left p = 0.62). However, term infants with visual pathway abnormality showed significantly higher VEP score than term infants without visual pathway abnormality (right p = 0.00, left: p = 0.00). In cases with visual pathway abnormality, comparing scores between term infants and preterm infants, term infants showed significantly lower score than the preterm infants in VEP study, at least for the right eye (right p = 0.046, left p = 0.27). In cases where there was no visual pathway abnormality, the term infants showed significantly lower score than the premature infants in both left and right eyes. Conclusions In term infants, we find more severe abnormal results in VEP study with visual pathway abnormality than without visual pathway abnormality. However, preterm infants may show severe VEP abnormality regardless of whether the visual pathway is abnormal or not. Follow up VEP studies may be of greater importance for preterm infants, regardless of the presence or absence of visual pathway anomaly, as the VEP abnormalities may be associated with brain immaturity in preterm infants.

5- Two types of acute zonal occult outer retinopathy differentiated by dark- and light-adapted perimetry K. Kazuki, H. Sakuramoto, Y. Nakao, C. Matsumoto, Y. Shimomura Department of Ophthalmology, Kinki University Faculty of Medicine, Osaka, Japan Purpose To assess the results of perimetry recorded under dark- and light-adapted (DA and LA) conditions in patients with acute zonal occult outer retinopathy (AZOOR) and to compare the results of electroretinography (ERG) and spectraldomain optical coherence tomography (SD-OCT) in two groups of AZOOR patients. Methods Twelve patients with AZOOR were studied. The diagnosis of AZOOR was based on the results of ophthalmoscopy, Goldmann kinetic perimetry, and multifocal ERGs (mfERGs). In addition, DA and LA perimetry, ERG, and SD-OCT were performed. The patients were followed for 1–9 years. Results The patients were classified into two types: Type A patients (3) had a scotoma detected by both DA and LA perimetry, normal or equally abnormal cone and rod ERGs, atrophy of the outer nuclear layer (ONL), and disruption of the inner segment/outer segment (IS/OS) junction line in the OCT images. Type B patients (7) had a scotoma that was more prominent in LA than in DA perimetry and a continuous IS/OS junction line in the OCT images. Two patients had characteristics of both type A and type B AZOOR. Conclusions Our findings suggest that eyes with type A AZOOR have focal and severe impairment of both the rods and cones, and eyes with type B AZOOR have focal and specific impairment of the cones.

Doc Ophthalmol (2014) 129:13–55 6- Flicker ERG recording in premature babies H. Sakai1, H. Minoda1, M. Yoshikawa2, E. Watanabe1, T. Tsuboi1, C. S. Matsumoto1,3, K. Shinoda1, A Mizota1 1 Department of Ophthalmology, Teikyo University School of Medicine, Tokyo; 2Mayo Corporations, Inazawa; 3 Matsumoto Eye Clinic, Tokushima, Japan

Purpose To record electroretinogram (ERG) in premature babies. Method: 28-Hz flicker ERG was recorded using a hand held ERG recording instrument (RETeval, Mayo, Inazawa, Japan) from both eyes of two premature babies; one was male, 1,902 g, 36 weeks of postconceptional age, 19 days old and the other was male, 1,928 g, 37 weeks of postconceptional age, 16 days old. Ohthalmoscopy revealed no retinopathy. Responses were collected with pupils fully dilated (Tropicamide 1 %) under room light using sensor strip skin electrodes that carried positive, negative, and reference electrodes which were attached on the lower lid. A mini Ganzfeld dome was placed in front of the eye and 3 cd s/m2 flash with 30 cd/m2 background light (ISCEV standard) was delivered to elicit the ERG response. Results The physiological response could be quickly and safely obtained. No systemic or topical adverse effect was observed. Conclusion Electrophysiological retinal function could be safely recorded in premature babies. Normative data collection and further study for repeatability and interpretation of each parameter is needed. 7- A case of optic atrophy with negative ERGs H. Yamazaki, Y. Chai Department of Ophthalmology, Kohnodai Hosp, National Center for Global Health and Medicine, City??, Japan Purpose We present detailed ophthalmic findings in a case of optic atrophy with negative ERGs. Methods A comprehensive ophthalmic examination, including full-field ERG and multifocal ERG, was performed. Results Present illness and family history: A 54-year-old man visited our hospital for decreasing visual acuity. He noted decreased visual acuity in his third decade of life and visual acuity loss progressed during recent years. His mother, his sister, and sister’s daughter reported having very poor visual acuity, but detailed ophthalmic examinations have not been done. Ophthalmic examinations His corrected visual acuity was 0.09 in both eyes with mild myopia. No abnormal findings were observed in the anterior chamber and vitreous. Fundus findings revealed optic atrophy in both eyes. No pigmentation was found in the periphery. Fluorescein angiography showed mild mottling. A thinning of the ganglion cell complex was detected by optical coherence tomography in both eyes. Goldmann kinetic perimetry showed pericentral scotoma and a peripheral defect in the nasal quadrant in the right eye, and enlarged blind spot in the left eye. Color vision was defective without a definite axis on the Panel D-15. Goldmann Weekers dark adaptometry showed a 2 log unit elevation of the final rod threshold. ISCEV protocol full-field ERG with an intense white

49 light stimulus after 30 min dark adaptation showed a normal a-wave amplitude but reduced b-wave amplitude (negative ERG). Amplitude of the b-wave of the cone ERG and 30-Hz flicker ERG were within normal range in both eyes. Amplitudes of the ON-responses for 150 ms duration stimuli were slightly reduced. Amplitudes of the OFF-responses were within normal range. Multifocal ERG showed decreased responses in the central area, but responses were preserved in the periphery. Conclusions The most common diagnoses associated with a negative ERG are X-linked juvenile retinoschisis, congenital stationary night blindness, central retinal artery occlusion, retinal toxicity from quinine and vincristin, and melanomaassociated retinopathy (MAR). In 1992, Weleber and Miyake reported two families with optic atrophy, negative ERGs, and relatively preserved full-field cone ERGs. They also reported focal macular ERGs that were essentially normal in their cases. Fundus findings and ERG findings in our case seem to be similar to those reported by Weleber and Miyake, except that detailed ophthalmic examinations of other family members have not been done. Our case showed normal full-field cone ERGs, but reduced amplitudes of multifocal ERGs in the central area. These findings may suggest that the case of optic atrophy with negative ERGs could be accompanied with the central cone dystrophy. Multifocal ERG is helpful in investigating the mechanism of this disease.

Toxicology, treatment 8- Effect of sodium valproate on the Retina Measured by Electroretinography M. Naser, S. M. Shushtarian Tehran Medical Branch, Islamic Azad University, Tehran, Iran Purpose Sodium Valproate or Depakine is a drug which is used in patients suffering from epilepsy. It is observed that some patients using Sodium Valproate complain of visual disturbances. The present study was designed to detect possible adverse effects of Sodium Valproate on the retina using Electroretinograghy (ERG) Method Twenty epileptic patients using Sodium Valproate were selected randomly. The patients were examined by routine ophthalmologic techniques to check for possible structural changes in the visual system. All subjects were selected with normal visual examinations using standard techniques. ERGs were recorded in all subjects. Peak latency (ms) and amplitude (lV) of the b-wave was measured. Results were compared with a control group. SPSS software was used to evaluate significant differences between the two groups. Results The b-wave peak latency was 44.2 ± 2.66 and 44.8 ± 2.36 ms in the case and control groups, respectively. The b-wave amplitude was 99.88 ± 12.18 and 99.16 ± 13.37 lV in case and control groups, respectively. The differences between two groups were not statistically significant. However, reviewing the history of individual patients we learned of some retinal problems in three subjects prior to Sodium Valproate exposure.

123

50 Conclusion From the result of present work one can conclude that Sodium Valproate does not affect retinal function as measured by the ERG. However, it is advised to examine the retina prior to Sodium Valproate prescription and to advise the patients of the safety of Sodium Valproate. 9- Multifocal electroretinography (mfERG) findings in patients with long-term hydroxychloroquine N. Agange, G. Abedi, J. M. Harrison University of Texas Health Science Center at San Antonio, Department of Ophthalmology, San Antonio, TX USA Purpose This is a retrospective chart review of patients on plaquenil who had mfERG. The objective is to study patients with abnormal mfERG on plaquenil to detect differences in mfERG pattern changes in patients for plaquenil toxicity compared to controls. Methods Institutional IRB approval was obtained prior to the study. Two-hundred and fifty mfERG reports from patients receiving plaquenil were reviewed and 91 with abnormal mfERGs were selected. Of the ninety-one patients, thirteen sets of eyes met the inclusion criteria. Data collected included other medical diagnoses, number of years on plaquenil, age, gender, BMI, cumulative dose, baseline visual acuity, Humphrey visual field (HVF) 10–2, mfERG with 103 hexagons. Ring 1 through 6 and ratios (1:2; 1:3; 1:4; 1:6; 2:6) were entered into an Excel database. The control group was comprised of patients on plaquenil with normal mfERG, no evidence of toxicity on biomicroscopy fundus examination, normal HVF, and normal visual acuities were used. High definition OCT was not available at the time of patient evaluations and follow up. Results The average age of study population was 56 years with females comprising 78 %. Systemic lupus erythematosus (SLE) made up 77 % of the population and 39 % of the subjects were on plaquenil for more than 10 years. The average cumulative dose was 786 g (SD: 921, CI 229–1,342). The average BMI was 28 (SD: 7.2; CI 24–33). The average baseline logMAR visual acuity of the right eye was 0.06 and 0.02 for the left eye. The average logMAR visual acuity at time of suspected toxicity in the right eye was 0.06 and 0.05 in the left eye, and were not significantly different from baseline (p = 0.97 and 0.58, respectively). At the time of suspected toxicity, only one subject had abnormal HVF 10–2 and one had detected maculopathy on biomicroscopy exam. However, all thirteen sets of eyes had at least one abnormal ring amplitude on mfERG at the time of suspected toxicity. The rate of abnormal values for ring amplitude and ratio in the right and left eye were 63, 74, 15, and 12 %, respectively. All these values were significantly different from 0, p \ 0.0001. Conclusions Our data shows that among the ring amplitudes, ring 1 seems to be the most sensitive for detecting early toxicity; however, this needs to be investigated further.

10- Snapshot of a hydroxychloroquine mfERG clinic M. Dolliver, E. Chelva, J. De Roach, S. Laurin, T. McLaren Department of Medical Technology & Physics, Sir Charles Gairdner Hospital, City, Australia

123

Doc Ophthalmol (2014) 129:13–55 Purpose An overview of the patient flow through an mfERG hydroxychloroquine clinic to manage increasing work flow and improve the manner of reporting for best patient outcomes. Methods This retrospective observational study examined records from patients attending the state referral visual electrophysiology clinic from November 2011 until January 2014. The hydroxychloroquine history and any possible renal or hepatic impairment were reported by the patient at the visit and rechecked at each subsequent visit. The mfERG data was recorded on the VERIS 6.3.2 allowing for an ICS (Index of Central Sparing) calculation. The mfERG testing complied with ISCEV guidelines and was referenced against age related normal ranges. Results Records were examined from 518 patients over 789 visits. These patients consisted of 86 % females aged 54 ± 14.6 years and 14 % males aged 59 ± 13.5 years. The diagnosis leading to hydroxychloroquine therapy was reported as rheumatoid arthritis in 41 % of patients, systemic lupus erythematosus in 38 % of patients, mixed connective tissue disorders in 6 % of patients and Sjogren syndrome in 5 % of patients; with the remaining 9 % recorded as various other autoimmune/inflammatory diseases. The mfERG results were subjectively categorized as normal/marginal (56 %), mildly abnormal (20 %) and abnormal (24 %). In our patient group, 39 % of patients were C60 years old, 36 % of patients reported hydroxychloroquine therapy for C10 years, 37 % of patients reported a cumulative dose C1,000 g, 24 % of patients’ calculated dosage per day was C6.5 mg/kg and 8 % of patients reported renal or hepatic impairment. ICS values were plotted against both duration of therapy and cumulative dosage. No correlation between an ICS value of 1 (suggestive of peri-central loss characteristic of hydroxychloroquine toxicity) and increased cumulative dose or duration was observed. This was true for all 1,525 visits (including right and left eye) and for a randomized sample (of either right or left) of 501 most recent visits. Recommendations by reporting ophthalmologists based on mfERG testing identified 48 patients in 2013 that should strongly consider ceasing hydroxychloroquine therapy. Conclusions Hydroxychloroquine toxicity is complex in nature and not easily predicted. There was no evidence of a strong relationship between an ICS value of 1 and dose regimen. In the best interest of our patients, a baseline measurement is essential with regular ongoing screening. Further investigation needs to be done into the role that autoimmune disease itself plays in changes seen on mfERG testing. 11- A case of ophthalmic artery occlusion following injection of hyaluronic acid into the glabellar area Y. Kitamura, T. Oshitari, S. Nonomura, G. Miura, A Chiba, S. Yamamoto Department of Ophthalmology and Visual Science, Chiba University Graduate School of Medicine Chiba City, Chiba, Japan Purpose To report a case of unilateral blindness due to ophthalmic artery occlusion following injection of hyaluronic acid into the glabellar area for facial soft-tissue augmentation. Methods A 20-year old woman underwent injection of hyaluronic acid into the glabellar area at an aesthetic plastic

Doc Ophthalmol (2014) 129:13–55 clinic. Immediately after injection, she suffered nausea, pain, paralysis of limbs, and visual loss in her right eye. Hyaluronidase was urgently injected into the same place, but the symptoms were not improved. She was emergently transported to our hospital. Results At the initial examination, the patient had no light perception in the right eye and her right pupil was dilated. The fundus examination revealed right central artery occlusion. No significant findings were detected on radiologic imaging of the head. Eye ball massage, hyperbaric oxygen therapy, and drip infusion of urokinase were conducted, but no improvement was seen. Fluorescein and indocyanine green angiography showed no filling of the right retinal and choroidal arteries. Electroretinogram in the right eye demostrated an absence of both a- and b-waves. Magnetic resonance angiography could not reveal the obstruction point of the ophthalmic artery. We considered that ophthalmic artery occlusion accompanied injection of hyaluronic acid into the glabellar area. Conclusions Although it is extremely rare, ophthalmologists and plastic surgeons should remember that an ophthalmic artery occlusion can develop in patients undergoing injection of hyaluronic acid. Our case illustrates that electroretinogram is helpful in differentiating CRAO from ophthalmic artery occlusion.

12- Assessment of retinal function just after intravitreal injection of an anti-VEGF agent G. Terauchi1, H. Minoda1, K. Sasaki1, M. Yoshikawa2, C. S. Matsumoto1,3, K. Shinoda1, A Mizota1 1

Teikyo University School of Medicine, Tokyo; 2Mayo Corporation, Inazawa; 3Matsumoto Eye Clinic. Tokushima, Japan Purpose To evaluate electrophysiological retinal function immediately after intravitreal injection of an anti-VEGF agent. Method Twenty-eight Hz flicker ERG was recorded using a hand held ERG recording instrument (RETeval, Mayo, Inazawa, Japan) before and after intravitreal injection of Bevacizumab (IVB, n = 6), Ranibizumab (IVR, n = 3), or Aflibercept (IVA, n = 6). Patients comprised of 11 males and four females and mean age was 69 ± 14 (mean ± SD) years old. Responses were collected with pupils fully dilated (Tropicamide 1 %) under room light. Sensor strip skin electrodes that carried positive, negative, and reference electrodes were attached on the lower lid. A mini Ganzfeld dome was placed in front of the eye and 3 cd s/m2 flash with 30 cd/m2 background light (ISCEV standard) was delivered to elicit responses. Results Post-operative ERG recording was performed 19–874 (167.1 ± 211.2) minutes after injection. The amplitude was 30.4 ± 3.18 and 30.7 ± 2.66 lV before and after injection, respectively. The peak latency was 6.08 ± 3.05 and 5.51 ± 3.11 ms before and after injection, respectively. No significant difference was observed in either parameter before and after injection. Conclusion Electrophysiological retinal function could be easily assessed before and after intravitreal injection, which

51 showed no harmful effect of the injection, at least in short-term after injection.

Inherited Retinal Disease 13- A cone dystrophy patient with a homozygous RP1L1 mutation S. Kameya1, S. Kikuchi1, K. Gocho1, S. El Shamieh2, K. Akeo1, Y. Sugawara3, K. Yamaki1, C. Zeitz2, I. Audo2, H. Takahashi4 1

Department of Ophthalmology, Nippon Medical School Chiba Hokusoh Hospital, Inzai, Chiba, Japan; 2INSERM, CNRS, UPMC Univ Paris 06, Department of Genetics, Institut de la Vision, Paris, France; 3Honjo Daiichi hospital, Yuri-Honjo, Akita, Japan; 4Department of Ophthalmology, Nippon Medical School, Bunkyo-ku, Tokyo, Japan

Purpose To describe a family with a cone dystrophy carrying a homozygous mutation in the RP1-like protein 1 (RP1L1) gene. Methods A family including one subject affected with cone dystrophy and three unaffected members underwent detailed ophthalmic and clinical evaluations including high-resolution imaging of retinal morphology by adaptive optics (AO) fundus camera. Mutation screening of the coding sequence of RP1L1 was performed by DNA sequencing analysis. Five computational assessment tools were used to predict the protein function (SIFT, Polyphen2, Align GVGD, PMut). In order to investigate other putative pathogenic variant(s), a nextgeneration-sequencing (NGS) approach was applied to the cone dystrophy patient with an exon sequencing array targeting 121 known genes underlying retinal diseases. Results The patient showed a mild reduction of cone and flicker responses on the full-field electroretinogram (ERG). Ellipsoid and interdigitation zone on spectral-domain optical coherence tomography (SD-OCT) images were severely disturbed. Focal macular ERGs showed a reduced amplitude of the a- and b-waves. Multifocal ERGs (mfERGs) responses were severely reduced. AO images of the patient showed severe reduction of cone density and an irregular cone mosaic. A homozygous RP1L1 mutation (c.3628 T[C) was identified in the patient. The mutation c.3628 T[C in exon 4 resulted in the substitution of proline for serine at amino acid position 1210 (p.S1210P). The serine at position 1210 is well conserved among the RP1L1 family in other species. Four out of five computational assessment tools predicted that this mutation is damaging to the protein function. This mutation was not present in 460 Japanese control alleles. Family members with heterozygous p.S1210P mutations showed normal best-corrected visual acuity, SD-OCT, mfERGs, focal macular ERGs, and AO analysis. The NGS approach only identified missense changes at the heterozygous state in PCDH15 (p.P923L), RPGRIP1 (p.V1211I), and GPR98 (p.L2422F), which did not co-segregate with the phenotype or/and were predicted to be benign. Conclusions We have demonstrated a possibility that autosomal recessive cone dystrophy may be caused by homozygous RP1L1 mutations.

123

52 14- Long-term follow-up of a large pedigree with a Phe211Leu mutation of the peripherin/RDS gene M. Andersson Gro¨nlund1,2, L. Holmegaard2, L. Tranebjærg3,4, J. Nilsson5 1

Institute of Neuroscience and Physiology/Ophthalmology, The Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden; 2Department of Pediatric Ophthalmology, The Queen Silvia Children’s Hospital at Sahlgrenska University Hospital, Gothenburg, Sweden; 3 Department of Audiology, H:S Bispebjerg Hospital, Copenhagen, Denmark; 4Wilhelm Johannsen Centre of Functional Genomics, Institute of Cellular and Molecular Medicine, ICMM, University of Copenhagen, Denamark; 5 Department of Clinical Neurophysiology, Institute of Neuroscience and Physiology, The Sahlgrenska Academy at the University of Gothenburg, Sweden Purpose To study the long-term clinical and electrophysiological progression, as well as the interfamily phenotype variability, for a peripherin/RDS Phe211Leu mutation associated retinal degeneration. Methods A detailed ophthalmologic evaluation including visual acuity (VA), refraction, kinetic visual field testing, slit-lamp inspection, ophthalmoscopy, dark adaptation final thresholds and a full-field ERG following ISCEV standards was performed. History of vision and hearing problems was taken and copies of previous medical records collected. Blood samples were drawn for genetic testing. Results were compared with previous testing (mean follow-up time 12.6 years, range 12–15 years). Previous full-field ERGs were recorded using Burian-Allen bipolar contact lens and the Nicolet Analysis System (Nicolet Biomedical Instruments, Madison, Wisconsin, USA). For the follow-up study, full-field ERGs were recorded with the Espion E2 system (Diagnosys LLC, Cambridge, UK). Prior to testing, pupils were fully dilated using cyclopentolate (0.85 %) + phenylephrine (1.5 %) and DTL-electrodes were placed in both eyes after administration of 1 % tetracaine drops. A ground electrode was attached to the forehead. Scotopic steps were recorded after a minimum of 30 min dark adaptation. Photopic steps were recorded after 10 min of light adaptation on a white background of 30 photopic cd m-2. Results Eleven affected subjects (seven male (M), four female (F) with a mean age of 53.4 years (range 24–76) at follow-up) and two non-affected siblings (one M, 58 years; one F, 50 years) were examined and included in the study. All affected subjects were positive for the p.F211L/N mutation. Most subjects followed a similar clinical course but one affected subject demonstrated a clearly different phenotype with slower progression. ERG responses were generally non recordable from age 40 years. On previous testing, ERGs were performed on nine affected individuals and in four of these, all responses were non-recordable above the noise level. The remaining five subjects had a pattern of rod-cone dysfunction with scotopic responses being non-recordable or greatly diminished and photopic responses diminished. At follow-up, four of five subjects with previously recordable ERG responses were found to have no recordable responses. One female (42 years) still had a pattern of rod-cone dysfunction with diminished but recordable mixed and photopic responses. In

123

Doc Ophthalmol (2014) 129:13–55 addition to the previously known affected subjects, two subjects (one M, 28 years; one F, 24 years) from a younger generation participated and were found to be affected. Five (four affected) out of the 13 subjects had hearing problems. Conclusions A Peripherin/RDS Phe211Leu mutation gives rise to a retinal degeneration with a variable phenotype and high interfamilial variability. Fundus changes and visual field constriction is present from at least early twenties. From approximately age 40 years, macular atrophy is evident and ERG responses are generally non-recordable. Best corrected VA remained above logMAR 0.5 in at least one eye in half of affected subjects over 70 years of age.

15- A phenotype-genotype correlation study in sib-ships with ABCA4-retinopathy M. Cotesta, J. Locke, R. Klatt, C. Westall, E. Heon, A. Vincent Department of Ophthalmology and Vision Sciences, The Hospital For Sick Children, Toronto, Canada Purpose To compare the clinical phenotype and electroretinographic parameters amongst seven sibling pairs with ABCA4related Stargardt’s disease (STGD). Methods A retrospective case series examined seven sib-ships with molecularly confirmed ABCA4-retinopathy. All cases underwent full-field electroretinogram (ERG) incorporating ISCEV standards. Details of clinical evaluation included best corrected visual acuity (BCVA), color vision and contrast sensitivity assessment, and fundus photography. Spectraldomain optical coherence tomography (SD-OCT) and fundus autofluorescence (FAF) results were obtained when available. Each subject was classified into one of the three groups based on electrophysiological findings: Group-1 (normal ERG), Group-2 (abnormal cone response and normal rod response) and Group-3 (abnormal rod and cone responses). Results All sib-ships were confirmed to have two disease causing variants in the ABCA4 gene. The ERG results were concordant amongst five sibling pairs; three sib-ships had normal ERGs (Group-1) and two sib-ships had abnormality in both rod and cone responses (Group-3). The four cases (two sib-ships) in Group-3 had cones more severely affected than rods (cone-rod pattern). Amongst the seven sib-ships, color vision results were concordant in all pairs except one; the discordant pair also demonstrated differences on BCVA ([2 lines). Group-3 sib-ships showed severe red-green and blueyellow deficits. The SD-OCT showed central retinal thinning with disruption of the external limiting membrane, and photoreceptor inner segment-outer segment junction in all tested cases. The FAF in the posterior pole showed generalized increase with peri-papillary preservation in all cases; the flecks were either hyper intense or atrophic and macular atrophy was observed. Conclusions The majority of sib-ships (71 %) showed good electrophysiological concordance. In the literature, ABCA4related retinopathy shows wide genotype-phenotype variation. The similarity in alleles amongst sib-ships across the whole genome is the probable reason for the phenotypic concordance observed in this study.

Doc Ophthalmol (2014) 129:13–55 16- Survival of the fittest in end-stage retinitis pigmentosa (RP): are OPs the last to die? A. L. Dorfman1*, M. Gauvin1*, J. M. Little1, R. K. Koenekoop1, J. -M. Lina2-3, P. Lachapelle1 * These authors equally contributed to this work 1

Department of Ophthalmology & NeurologyNeurosurgery, McGill University, Montreal Children’s Hospital Research Institute,; 2De´partement de ge´nie e´lectrique, E´cole de Technologie Supe´rieur,; 3Centre de Recherches Mathe´matiques, Montre´al, Que´bec, Canada

Purpose Current knowledge of how the ERG signal is generated suggests that it consists of slow potentials, namely the a- and b-wave, onto which the faster OPs are superposed. Previous studies have shown that the OPs alone are affected in some disease processes, while the a- and b-wave remain relatively intact (such as in diabetic retinopathy), therefore suggesting that the OPs are produced independently from the b-wave. The purpose of this study was to examine whether a signal solely composed of OPs could be generated with certain pathologies. Methods A retrospective analysis of photopic ERGs (at the peak of the photopic hill, flash intensity: 0.64 log cd s m2; background: 30 cd m2) was carried out while paying particular attention to those that had been previously reported by us to be ‘‘non-recordable’’, or residual responses. A visual inspection of the ERG was performed using a template of a normal photopic OP response which was superposed on the responses obtained from our patients to see if there was a wave match. Those that were indeed found to have a wave match were subsequently analyzed using the discrete wavelet transform (DWT) to quantify the contribution of the OP frequency domain to the ERG [%OP = OP band/Total ERG band = (80–160 Hz band energy/20–160 Hz band energy)]. Findings were compared to the %OP calculated from normal patients. Results Of all the patients considered (&750), only 10 qualified. Those patients were reported to be at the end-stage of RP and were characterized with visual fields B10 (both eyes), visual acuities ranging between 20/25 and light perception, and nonrecordable rod ERG responses. While in normal subjects the OPs accounted for 35 % ± 3 of the total energy contained in the photopic responses, in pathological cases they accounted for more than 68 % ± 16 (min: 40 %; max: 91 %); p \ 0.05. Conclusion Progressive retinal degenerations (such as RP, for example) are often characterized by a significant reduction in the visual field, which ultimately results in tunnel vision by the end-stage of the disease. Although we do not know exactly how the b-wave is generated, it must require a balance between the retinal interneurons (such as bipolar cells) and the Muller cells. Should an imbalance occur, such as might be the case in advanced RP where only a minimal amount of healthy retina remains, one wonders whether the bipolar to Muller cell ratio increases, with sufficient bipolar cells available to generate the OPs, yet fewer Muller cells to generate the b-wave. Consequently, as the visual field constricts, the ratio of the OPs over the b-wave would gradually increase, to the point where only OPs would remain, a concept to be verified in patients with more recordable ERGs.

53 Funding This project was supported by CIHR and FRQ-S, under the frame of E-Rare-2, the ERA-Net for Research on Rare Diseases.

17- Novel funduscopic features in Oguchi’s disease; relationship between retinal arteries and golden metallic reflex Y. Kato1,2, K. Tsunoda1, K. Fujinami1,3,4, T. Noda1, Y. Oguchi3,5 1

Laboratory of Visual Physiology, National Institute of Sensory Organs, Tokyo Medical Center, Tokyo, Japan; 2 Department of Ophthalmology, Tokyo Metropolitan Komagome Hospital, Tokyo, Japan; 3Department of Ophthalmology, Keio University School of Medicine, Tokyo, Japan; 4UCL Institute of Ophthalmology, London, UK; 5Department of Ophthalmology, Mizumachi Clinic, Tokyo, Japan Purpose Oguchi’s disease is a form of congenital night blindness characterized by golden metallic reflex of the fundus in the light-adapted condition (Oguchi 1907). The origin of this peculiar reflex has been investigated by various approaches; histopathological assessment (Yamanaka 1924 and 1969, Kuwabara 1963), intra-retinal injection of potassium chloride (de Jong 1991), surgical removal of the vitreous (Kuroda 2011), scanning laser ophthalmoscopy (Usui 2000) and optical coherence tomography (OCT) (Yamada 2009, Takada 2011). The underlying mechanism, however, has not been clearly identified. By carefully observing the fundi in patients with Oguchi’s disease, we have found novel funduscopic features, which had not been reported in the literature. Methods Detailed ophthalmological examinations were performed in 11 cases with Oguchi’s disease (four males and seven females, age 10–75). In all the cases, the fundus showed diffuse or localized golden metallic reflex, and fullfield ERGs demonstrated typical findings compatible with Oguchi’s disease; undetectable rod response after 30 min dark adaptation, electronegative configuration of mixed rod-cone response with reduced amplitude of the a-wave, and normal cone responses. Genetic screening was performed in seven cases and all of them had the same homozygous mutation in the SAG gene (c.926delA), which had been frequently found in Japanese patients with Oguchi’s disease. The retinal appearances from the posterior pole to the equator could be confirmed by photographs in twenty eyes of eleven cases. In two cases with SAG mutation (Case 1; 18-year-old male and Case 2; 23-year-old female), fundus photographs had been serially obtained for 6 years in order to observe the alteration of the metallic reflex. Results The metallic reflex was diffusely observed only in three out of 20 eyes, whereas the reflex was not homogeneously distributed in the other eyes. In 11 eyes of seven cases, there were clearly demarcated regions without metallic reflex. In seven eyes, their border was located along retinal arteries, but not veins. Prolonged dark adaptation over 2.5 h diminished the metallic reflex (Mizuo-Nakamura phenomenon) (Mizuo 1914), and the demarcated regions became unclear or disappeared. In Cases 1 and 2, the discrete regions were extended along the retinal arteries during 6 years’ follow-up.

123

54 The high- or low-reflectivity regions along the peripheral retinal vessels were observed in all the cases, as reported by Oguchi (Oguchi 1907). It was notable that this funduscopic feature was observed only along arteries in all the cases except two. OCT demonstrated abnormally high reflectivity of the photoreceptor layer in the regions with metallic reflex. Conclusion Recent studies using the OCT have suggested that photoreceptor layer is strongly related to the occurrence of the metallic reflex in Oguchi’s disease. The contribution of retinal arteries observed in our cases does not conflict with this idea but raised another possibility that the metallic reflex is strongly influenced by the superficial structures of the retina, such as retinal artery, vitreous and vitreo-retinal interface. 18- Analysis of electroretinogram phenotype in congenital stationary night blindness J. Locke, M. Cotesta, R. Klatt, C. Westall, E. Heon, A. Vincent Department of Ophthalmology and Vision Sciences, The Hospital For Sick Children, Toronto, Canada Purpose To examine the full-field electroretinogram (ERG) variability in patients with molecularly confirmed CACNA1F, CABP4 and CACNA2D4-related disease—the three genes known to result in an incomplete type of Schubert Bornschein ERG abnormality in congenital stationary night blindness (CSNB). Methods This retrospective study reviewed ISCEV Standard ERG recordings from 19 patients (Age Range: 1–17 years) with molecularly confirmed CACNA1F, CABP4 or CACNA2D4 mutations. Dark adapted scotopic ERG (DA 0.002), mixed rod-cone ERG (DA 2.29), single flash photopic ERG (LA 2.29) and photopic 30 Hz flicker response (LA30 Hz 2.29) were reviewed. Amplitudes and implicit times of the a- and b-waves were recorded. The results were compared to age and electrode- matched control data (n = 41). Details of clinical examination were available in all cases. Results Sixteen cases had mutations in the CACNA1F gene (X-linked), two cases had mutations in the CABP4 gene (autosomal recessive) and one had mutations in CACNA2D4 gene (autosomal recessive). Eight patients had a myopic refractive error (all CACNA1F), and nine patients had a hyperopic refractive error (CACNA1F—six cases, CABP4— two cases, CACNA2D4—one case). Nine CACNA1F-related CSNB cases had electrophysiology phenotype consistent with incomplete CSNB; reduced but recordable DA 0.002 ERG b-wave, electronegative 2.29 ERG, reduced and/or delayed LA2.29 ERG b-wave, and reduced and delayed LA2.29 30 Hz b-wave. Of the remaining seven CACNA1F cases, DA0.002 ERG was either normal (n = 1) or non-recordable (n = 6); all except the one with normal DA0.002 b-wave had electronegative DA2.29 ERG. LA ERGs in all seven cases were consistent with incomplete CSNB. Both cases with CABP4 mutations showed abnormal LA ERGs; but only one of them showed electronegative configuration in DA2.29 ERG. The lone case with CACNA2D4 mutation showed normal DA ERGs; LA ERGs showed reduced amplitudes and delayed implicit times.

123

Doc Ophthalmol (2014) 129:13–55 Conclusions The photopic ERG (LA 2.29 2 Hz and 30 Hz) abnormalities consistent with incomplete CSNB were observed in all cases. Electronegative configuration of DA2.29 ERG was noted in the majority (84 %), but may occasionally be normal. DA0.002 ERG b-wave showed most variability as it could subnormal (53 %), non-recordable (37 %) or normal (10 %). The genotype could not explain the phenotype variability in DA ERGs; both cases that had normal DA0.002 ERGs had severe mutations, either non-sense or gross intra-genic deletions.

19- Clinical and electrophysiological phenotype in autosomal dominant pattern dystrophy A. Vincent, C. Westall, E. Heon Department of Ophthalmology and Vision Sciences, Hospital for Sick Children, University of Toronto Toronto, Canada Purpose To report the clinical, electrophysiological and genetic results in autosomal dominant pattern dystrophy (ADPD). Methods Four probands with ADPD underwent detailed clinical evaluation including best corrected visual acuity (BCVA) testing, color vision testing, contrast sensitivity measurement and fundus photography. Spectral-domain optical coherence tomography (SD-OCT) and fundus autofluorescence (FAF) testing were available in all four probands. ISCEV Standard ERGs and sensory electro-oculogram (EOG) was available in three and two probands respectively. Additional affected family members of two of the probands also underwent detailed clinical and electrophysiological testing. Molecular genetic testing was performed in all four probands. Results The age of the probands ranged between 8 and 68 years. BCVA was better than 20/60 in all except the left eye of Case 3 (46 years), in which there was a macular scar. Color vision testing showed mild red-green and blue-yellow deficits only in two probands. Fundus evaluation demonstrated a butterfly pattern (n = 2), multifocal pattern dystrophy simulating Stargardt disease (n = 1) or grouped pigmentary pattern (n = 1). The SD-OCT showed localized areas of RPE deposition in all probands. ISCEV standard ERGs were borderline normal in two patients. Case 4 (8 years) showed reduction in both rod and cone response amplitudes. The Arden ratio was normal in one patient (Case 2; Age 68 years) and borderline in another (Case 3; 46 years). The pattern at the macula showed intra-familial variability in additional examined family members. Heterozygous mutations in RDS/PRPH2 were identified only in two probands; BEST1 screening returned negative in the other two. Conclusions ADPD is a phenotypically heterogeneous disorder that demonstrates wide inter- and intra-familial variability. Central vision impairment (BCVA, color and contrast sensitivity) is not severe unless there is macular scarring secondary to choroidal neo-vascular membrane formation. The full-field ERG may be subnormal in a proportion of cases, suggestive of a generalized retinal dysfunction. The disorder appears to be genetically heterogenous as mutations in RDS/PRPH2 or BEST1 are not found in all cases of ADPD.

Doc Ophthalmol (2014) 129:13–55 20- Wagner vitreoretinopathy mimicking retinitis pigmentosa H. Li1, H. Wang2, R. Chen2, H. Li1, R. Su1 1 Department of Ophthalmology, Peking Union Medical College Hospital, Peking Union Medical College, Beijing, China; 2Human Genome Sequencing Center, Baylor College of Medicine, Houston, Texas, USA

Purpose To study the phenotype and genotype features of a 3-generation Chinese family with retinal dystrophy mimicking retinitis pigmentosa. This was eventually re-diagnosed as Wagner vitreoretinopathy. This was based on targeted DNA capture combined with next-generation sequencing (NGS). Methods Seven subjects were involved in this study, including three affected and four unaffected individuals. The proband and her parents underwent detailed ophthalmic examinations. The DNA sample of the proband was sequenced using our customized capture panel, which includes 214 retinal disease

55 genes. Sanger sequencing was performed for segregation and confirmation. Results The affected subjects manifested decreased visual acuity, chorioretinal atrophy, pigment clumps and vitreous veils. Abnormal ERGs with moderately reduced b waves were recorded. NGS data revealed a novel splice mutation (c.6304 ? 2T [ A) in VCAN and a nonsynonymous variant in RP1. A segregation test showed all three patients carried this heterozygous VCAN splice site mutation. The proband and her normal father carried the heterozygous RP1 gene (p.L1130I), which is polymorphic. Conclusions A Chinese family with autosomal dominant retinal dystrophy was re-diagnosed as Wagner vitreoretinopathy after NGS. A novel splice mutation was identified. Our findings highlight the advantage of NGS in comprehensive molecular diagnosis of inherited retinopathy, especially those highly heterogeneous with overlapped phenotypes.

123