An analysis of observer‐rated functional vision in patients implanted with the Argus II Retinal Prosthesis System at three years

Abstract

Introduction

The purpose of a retinal prosthetic device is to replace the function of dead or dying photoreceptor cells, which cause visual loss. This is achieved by electrical stimulation of the remaining secondary neurons (for example, ganglion and bipolar cells) in such a way that the signals can be processed by the brain to generate visual perception.

Electrodes intended to stimulate secondary neurons can be implanted on, in or behind the retina, using an epiretinal, subretinal or suprachoroidal approach, respectively.1 In the subretinal approach, the implant is placed in the space between the pigment epithelial cells and the dead photoreceptors, whereas in the epiretinal approach, the implant is placed on the surface of the retina. The suprachoroidal electrode array is inserted between the sclera and the choroid.

One of the first clinical applications of retinal prostheses is the Argus II Retinal Prosthesis System (Second Sight Medical Products, Inc, Sylmar, California, USA), which was approved by the Food and Drug Administration (FDA or Agency) in 2013 for treatment of late‐stage retinitis pigmentosa. The device includes a small video camera that is mounted on eyeglass frames. The camera is connected to a processing device worn by the patient, which converts the video input to electronic signals that are transmitted wirelessly to an electrode array (60 electrodes arranged in a 6 × 10 grid) implanted on the surface of the retina (that is, epiretinal implant). The array uses this information to stimulate remaining healthy cells in the retina, and visual information is thereby transmitted by the optic nerve to the brain, where it is perceived as patterns of light. The patient learns to interpret these light patterns.

Safety and benefit of the Argus II System were established from a prospective 30‐patient single‐arm clinical study, which has been previously published.2, 3 Patients participating in the study had baseline visual acuity of bare light perception or no light perception (that is, 2.9 logMAR in both eyes [worse than 6/4,766 in Snellen notation]), with the device implanted in the worse‐seeing eye. Benefit was established from a suite of assessments, including three computer‐based, objective tests of basic visual skills, intended to measure incremental changes in a profound vision‐loss population. Outcomes demonstrated improvement in visual function, with 33 to 48 per cent of patients scoring better than 2.9 logMAR after three years and one year of use, respectively.

Just as objective tests of basic visual skills had to be developed to establish visual function benefits, standardised assessments are required to evaluate functional vision and quality of life benefits; however, in patients with profound visual loss (defined as 6/150 to 6/300 or logMAR 1.4 to 1.7)4 or worse, there are no validated assessments of functional vision. For example, both the National Eye Institute Visual Function Questionnaire (NEI‐VFQ‐25)5 and the Veteran Affairs Low Vision Visual Functioning Questionnaire (LV VFQ‐48),6 have relatively few items that can be completed by patients with profound vision loss and instead are most appropriate for use with patients with low vision, for example, better than 6/120 or logMAR 1.3.7, 8

Therefore, the Functional Low‐Vision Observer Rated Assessment (FLORA) instrument was developed specifically for use in patients implanted with retinal prostheses who suffer from profound loss of vision or blindness. The structure and development of FLORA has been previously published.9 The instrument combines a self‐report section (Part 1) with a list of functional visual tasks that are rated by a qualified evaluator (Part 2). The tasks were selected based on the requirement that vision be used primarily to achieve some aspect of the task (versus other senses) and that the vision provided by the prosthetic device had the potential to improve performance. For example, selected tasks require the use of light perception (such as, use light from windows to determine orientation), and movement, spatial or form vision (such as, recognise shapes, detect curbs, track another person). Activities requiring higher levels of vision such as reading 12‐point print and driving are not included, as the amount of vision provided by current retinal prostheses is not sufficient to complete these tasks visually. A case‐report narrative (Part 3) summarises the assessor's evaluation of the effect of the Argus II System on the subject's life.

Tasks are organised into four domains, including ‘Visual orientation’, ‘Mobility’, ‘Daily life’ and ‘Interaction with others’. Trained evaluators observed a patient performing each assessed task with the system ON and with it OFF. The ease with which a task is completed is categorised using a four‐point scale.

The purpose of this paper is to analyse observer‐rated functional vision using FLORA (Part 2) in patients implanted with the Argus II System, comparing outcomes with the device ON and OFF.

Methods

Flora

An assessment of functional visual abilities in and around a residential setting was measured using FLORA. FLORA is a multi‐part instrument primarily developed to obtain an observer‐rated assessment of how patients use a retinal prosthesis in completing a series of common activities of daily living. Thirty‐five tasks are organised into four domains: Visual orientation, Visual mobility, Daily life and Interaction with others. Each of the four domains is intended to measure a different aspect of functional vision. Visual orientation evaluates the ability to use light projection and contrast to improve spatial orientation (that is, understanding where and how one is positioned in the environment). Visual Mobility primarily measures the ability to use vision to detect obstacles. Daily living activities are a series of common activities typically done in a familiar or common environment, such as locating items in a bathroom and locating clothes. Finally, Social interaction measures how patients interact with others in a social setting, including detecting the approach of a person and determining the direction another person is facing.

All assessments were made by qualified evaluators. These evaluators observed a patient performing each assessed task with the system ON and with it OFF (a control condition). The ease with which a task is completed was categorised using a four‐point scale, ranging from easy (score of 1) to impossible (score of 4). In addition, evaluators estimated how much vision is used to accomplish the tasks (ranging from no vision, some vision or only vision).

All FLORA evaluations occurred at the patient's residence and local environment to measure outcomes that were directly relevant to the patients. No effort was made to control lighting (natural or artificial) or other environmental factors to ensure that the assessment was reflective of a ‘real‐world’ experience.

Evaluators were instructed to select tasks based upon each patient's self‐reported goals and activities. As a result, no patient completed all 35 tasks but rather a subset of tasks, which were not necessarily the same from patient to patient.

All evaluators were independent contractors certified in the areas of rehabilitation for the blind, low vision, orientation and mobility or occupational therapy and completed company‐sponsored training on the implementation and use of FLORA. Depending on the centre, one or two therapists implemented the FLORA. When two therapists performed a FLORA for a particular subject, they shared responsibilities; this was usually done with one occupational therapist and one orientation and mobility specialist, who divided the tasks according to their specialties (that is, duplicate assessments were not made). All evaluators were paid for time and out‐of‐pocket expenses. Per the request of the FDA, none of the evaluators were employees of Second Sight or had any financial stake in the Company.

Results

A total of 26 out of 30 eligible Argus II patients were assessed with the FLORA between 18 to 44 months (mean 36 months) after the implant. Data were not available on one patient due to the device being explanted after 14 months. An additional three patients did not consent to the FLORA and thus, were excluded from the analysis.

A list of all 35 tasks, sub‐divided by domain, and the number of patients completing each task (with the device both ON and OFF), is provided in Table 1. Overall, 24 (69 per cent) of the tasks showed a statistically significant improvement in outcome (that is, were easier to perform) with the device ON versus OFF (indicated with an asterisk in Table 1). Two tasks (six per cent) showed a statistically significant decrease in outcome with the device ON versus OFF (indicated with a double asterisk in Table 1). The remainder of tasks (nine or 26 per cent) showed no significant change between ON versus OFF.

For all tasks within the Orientation, Mobility and Interaction domains, there was a significant improvement in task‐completion score with the device ON versus OFF with the exception of tasks 5, 34 and 35 (Table 1: 5. Use the sun to determine orientation. 34. Visually locate people in a crowded setting and 35. Determine direction another person is facing).

Within the 16 items of the Daily Life domain, there was a significant improvement in task‐completion score for all activities with the exception of five tasks: 21 to 23 and 27 and 28. For two of the tasks (21 and 22: Travel within home independently and Identify top step/bottom step), the OFF data were significantly better than the ON data, whereas for tasks 23, 27 and 28, there was no significant difference between ON and OFF.

The largest improvement in ON versus OFF scores for individual activities was for Tasks 1, 2, 4, 12 and 20 (which ranged from −1.89 to −1.52). All five tasks have in common the ability to use light projection and contrast to identify objects.

The changes in average score of observer‐rated tasks by domain, when the device was ON versus OFF, are provided in Table 2. The ease in which a task is completed is assessed using a four‐point scale, ranging from easy (score of 1) to impossible (score of 4). Thus, a reduction in score reflects an improvement in task completion.

As is evident from Table 2, within each domain, ON values were lower (thus better) than OFF values. In each case, the comparison of ON versus OFF was statistically significant. Based on the percent change in score, the Orientation domain showed the largest improvement at 38 per cent with Daily life having the smallest improvement at 19 per cent.

Discussion

Conclusion

The FLORA was developed in accordance with FDA guidance to measure functional vision in a patient's own living and working environment. Part 2 of the instrument was used to measure the change in observer‐rated tasks in patients implanted with the Argus II Retinal Prosthesis System, when the device was ON versus OFF. In all domains and the majority of individual tasks, ON data were significantly better than OFF data, meaning that task completion was easier when patients used the Argus II System. Improvements ranged between 19 to 38 per cent within the four domains, with the greatest improvement measured in tasks involving light projection and contrast. These data demonstrate that within a population with profound visual loss treated with a retinal prosthesis, measuring functional vision is relevant to establishing efficacy and utility.

Disclosures

Dr Jessy Dorn and Dr Robert Greenberg are employees and stockholders of Second Sight Medical Products.

Notes

Geruschat, D. R. , Richards, T. P. , Arditi, A. , da Cruz, L. , Dagnelie, G. , Dorn, J. D. , Duncan, J. L. , Ho, A. C. , Olmos de Koo, L. C. , Sahel, J‐A. , Stanga, P. E. , Thumann, G. , Wang, V. , and Greenberg, R. J. (2016) An analysis of observer‐rated functional vision in patients implanted with the Argus II Retinal Prosthesis System at three years. Clin Exp Optom, 99: 227–232. doi: 10.1111/cxo.12359. [PMC free article] [PubMed] [Google Scholar]

References