June 2016
Volume 57, Issue 7
Open Access
Retina  |   June 2016
Automated Light- and Dark-Adapted Perimetry for Evaluating Retinitis Pigmentosa: Filling a Need to Accommodate Multicenter Clinical Trials
Author Affiliations & Notes
  • David B. McGuigan, III
    Scheie Eye Institute Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
  • Alejandro J. Roman
    Scheie Eye Institute Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
  • Artur V. Cideciyan
    Scheie Eye Institute Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
  • Rodrigo Matsui
    Scheie Eye Institute Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
  • Michaela L. Gruzensky
    Scheie Eye Institute Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
  • Rebecca Sheplock
    Scheie Eye Institute Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
  • Samuel G. Jacobson
    Scheie Eye Institute Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
  • Correspondence: Samuel G. Jacobson, Scheie Eye Institute, University of Pennsylvania, 51 N. 39th Street, Philadelphia, PA 19104, USA; jacobsos@mail.med.upenn.edu
Investigative Ophthalmology & Visual Science June 2016, Vol.57, 3118-3128. doi:10.1167/iovs.16-19302
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      David B. McGuigan, III, Alejandro J. Roman, Artur V. Cideciyan, Rodrigo Matsui, Michaela L. Gruzensky, Rebecca Sheplock, Samuel G. Jacobson; Automated Light- and Dark-Adapted Perimetry for Evaluating Retinitis Pigmentosa: Filling a Need to Accommodate Multicenter Clinical Trials. Invest. Ophthalmol. Vis. Sci. 2016;57(7):3118-3128. doi: 10.1167/iovs.16-19302.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: The purpose of this study was to develop a convenient means to measure rod (and cone) function by automated perimetry in patients with inherited retinal degenerations (IRDs).

Methods: A currently available automated perimeter was used to determine sensitivity (in decibels) to a blue target in the dark-adapted (DA) state and a white target in the light-adapted (LA) state. Normal subjects and IRD patients were evaluated with a full-threshold 71-locus strategy (the retinitis pigmentosa [RP] test) and a size III target. Comparisons were made with results from the more commonly used methods of two-color DA perimetry and middle/long-wavelength LA perimetry in the same patients.

Results: Rod function using the blue target and the RP test was determined for normal subjects by measuring DA sensitivities. If patients detected the blue stimulus in the DA state, it was determined whether the value was rod mediated by using normal data acquired during the cone plateau phase of dark adaptation. If rod mediated, rod sensitivity loss (RSL) was calculated and mapped across the visual field. Light-adapted sensitivities in normal subjects were also measured, permitting cone sensitivity losses (CSL) to be calculated for the patients. Multiple methods were used to compare RSL and CSL results with those from two-color DA perimetry and chromatic LA perimetry, and there was close correspondence between the methods.

Conclusions: The unmodified automated static perimeter used in the DA and LA states presents a practical approach to accomplish current goals of treatment trials in IRDs. This proof-of-principle study is an initial step toward establishing a clinical method to gather reproducible data on photoreceptor-mediated sensitivity.

The evaluation of patients with inherited retinal degenerations (IRDs) has evolved over decades from standard clinical eye examinations to an era when electroretinography (ERG) was the key test to understand phenotypes.1 Psychophysical tests to determine rod- and cone-based visual function in different retinal regions were developed to understand the different disease patterns across the retina otherwise not revealed by a retina-wide sum, such as the ERG.2 Most commonly such attempts have used static threshold perimetry, but there has also been an attempt to develop kinetic rod perimetry.3 Now we have cross-sectional and en face imaging to contribute to our understanding of these patients (for example, Refs. 4–8). The various methods used to evaluate phenotype took on a different importance when genotypes were discovered. A gene-based understanding of disease mechanisms of the many different human IRDs has been a major step forward, and phenotype in the context of genotype has helped decipher the complexities of the disorders.911 
The concept of measuring spectral sensitivity at various locations in the visual field to determine rod and cone function in IRD patients is longstanding.12 Measuring the regional variation of rod and cone dysfunction across larger extents of the visual field was made possible using two-color dark-adapted (DA) static perimetry, first performed manually.2 Automation of DA static perimetry with light-emitting diodes and a purpose-built instrument was a further advance,13,14 followed by modification of commercially-available computerized projection perimeters.15,16 Two-color DA perimetry permitted rod function to be quantified across the visual field, but, when there was substantial measurable rod function, L/M (long-middle wavelength) cone function was usually not easily isolated. Increment thresholds, on a conventional white background, were then used with white or middle/long wavelength stimuli to be able to acquire cone-based results to complement the DA function.17 
Many IRD patients have been characterized using DA automated static perimetry, and the technique has been used more recently in certain clinical trials.18,19 Given that we have entered an era of therapeutics, some of which are focal (e.g., subretinal gene delivery) or not expected to affect the entire retina (e.g., intravitreal delivery of some therapies20), there is a need to make DA perimetry a technique that can be implemented in most clinics rather than only in the rare clinic specializing in IRDs. In an attempt to meet this need, we reassessed the feasibility of DA automated perimetry with a commercially available instrument, designed protocols that could be used in most ophthalmology clinics, established normal limits for the protocols, examined a series of patients with IRDs for feasibility of technique, and compared the results with those from a published method.15 
Methods
Human Subjects
Eighteen IRD patients (age, 24–66 years) were included (Table). Patients had a complete eye examination. A group of normal subjects (n = 11; age, 22–57 years) were also studied. Informed consent was obtained; procedures followed the Declaration of Helsinki, and there was institutional review board approval. 
Table
 
Clinical Characteristics of the IRD Patients
Table
 
Clinical Characteristics of the IRD Patients
Localized Visual Sensitivities
Protocols determined rod- and cone-mediated visual sensitivities at 71 loci on a 12° grid across the visual field: extending from 72° temporally, 48° nasally, 36° superiorly, and 48° inferiorly. Foveal sensitivity was determined using a diamond-shaped array of light emitting diodes as the fixation target. This series of 71 loci will be referred to as the retinitis pigmentosa (RP) test. Two strategies were used, both with full-threshold testing: first, our previously published method,15,21 and second, the new method that is the topic of the current work. First, chromatic static perimetry (500- and 650-nm stimuli, DA; 600-nm, LA on a 10-cd/m2 white background; 200-ms duration, size V, 1.7°-diameter target) was performed with a modified computerized perimeter (Humphrey Field Analyzer [HFA], HFA-750i analyzer; Zeiss-Humphrey, Dublin, CA, USA) as described.15,21 Rod-mediated locations were determined by the sensitivity difference between 500- and 650-nm stimuli.15,21 Dark adaptation was accomplished with an opaque patch over the eye to be examined or using light-reducing ski goggles with red filters, if both eyes were being examined. 
The second method used the same computerized HFA perimeter (HFA-750i; Zeiss-Humphrey) but without modifications to the optical pathway of the instrument. The new method used the internal blue filter in the DA state. The blue filter is used conventionally in the “blue-on-yellow” condition (short-wavelength automated perimetry22). The target size was the conventional one used in glaucoma diagnosis and management (size III, 0.43° diameter). Only in the acquisition of DA results from normal subjects was it necessary to use an additional attenuation of the stimulus with a 1.5-log-unit neutral density filter to avoid the ceiling effect due to the higher range of normal values. 
The two methods were compared by 95% limits of agreement analysis for repeated measures with linked replications2325 using the R statistical software (MethComp package26). This R package is used for analysis of studies in which two methods of measurement are compared and is based on the lme (linear mixed-effects model) procedure.27 
Results
Determining Regional Rod- and Cone-Mediated Function With a Commercially Available Automated Perimeter
To assay the regional variation of rod function across the visual field with a method that could have widespread clinical use, we altered our previous strategy of two-color DA static perimetry.15 Our original method required modifications of the HFA that involved optical and mechanical engineering so that short- and long-wavelength interference filters and neutral density filters could be inserted into the optical pathway of the light stimulus. This has been both feasible and useful for characterizing the regional variation of rod- and cone-mediated function in IRD patients. As the HFA models advanced technologically, we revised our modifications to accommodate each new generation of instruments. 
The new method to assay rod function, however, uses an unmodified HFA, permitting DA perimetry to be performed on IRD patients in a clinic that would otherwise be using LA perimetry for patients with other eye diseases that require automated perimetry, such as glaucoma. Numerical DA blue sensitivity results for the RP test strategy in normal subjects (displayed as right eye data) are shown (Fig. 1A). Locus-by-locus mean results (Fig. 1A, left) and lower limits of normal (Fig. 1A, right) are given. In general, mean sensitivity to this blue target is relatively flat (∼45–55 dB) across the field sampled; sensitivity at fixation is lower. The results compare favorably with DA sensitivities to the short-wavelength target in normal subjects using the modified HFA and the larger target.15 To define for each locus the lowest sensitivity level that can be ascribed to rod-mediated function, in normal subjects, we determined DA blue sensitivity at cone plateau, that is, after a period of light adaptation known to desensitize rods and before rods recover (lower inset to the right of Fig. 1B). The mean sensitivity results for the RP test (blue stimulus, size III) at cone plateau in a subset of normal subjects are shown (Fig. 1B). The DA sensitivity of a patient is considered to be detected by rods if it falls above the lower limit of normal DA blue at cone plateau. 
Figure 1
 
Normal data for DA and LA RP tests using the unmodified HFA. (A) Dark-adapted blue normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD). (B) Dark-adapted blue at cone plateau. (Left) Mean normal. (Right) Dark adaptation functions indicating with brackets the concept of final DA thresholds (upper) and at cone plateau (lower). (C) Light-adapted white normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD).
Figure 1
 
Normal data for DA and LA RP tests using the unmodified HFA. (A) Dark-adapted blue normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD). (B) Dark-adapted blue at cone plateau. (Left) Mean normal. (Right) Dark adaptation functions indicating with brackets the concept of final DA thresholds (upper) and at cone plateau (lower). (C) Light-adapted white normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD).
For LA perimetry (10 cd/m2 white background light), the achromatic (white) stimulus is used with the size III target. This is a common strategy in glaucoma monitoring, and such increment thresholds with automated perimetry have also been used in clinical trials of IRDs.2831 For the present purpose, the specific test strategy would be the RP test so that the results are on the same grid as those from DA testing. Mean normal results across the RP test loci are shown (Fig. 1C, left), and the lower limits of normal at each locus are also given (Fig. 1C, right). 
Rod- and Cone-Mediated Function Sampled Across the Visual Field of Representative IRD Patients
Many patterns of peripheral visual field abnormality occur in IRDs.32 Once the sensitivities to the blue stimulus are recorded in the DA state in an IRD patient, questions about the results arise: Are there any rod-mediated loci? If so, is rod function normal or abnormally reduced? If reduced, by how much? The LA sensitivities to a white stimulus need less interpretation and only the relationship to normal results. 
Data from two IRD patients are shown as examples of DA and LA results and interpretations thereof (Figs. 2, 3). Patient 15 is a 56-year-old woman with autosomal dominant RP due to the G106R rhodopsin mutation. With DA perimetry (Fig. 2A), there were 29 extrafoveal loci detected with the blue stimulus, mainly in the inferior field. Of these loci, all but one locus meet our criterion for being rod-mediated. Rod sensitivity loss (RSL) is the difference between patient result at a locus and mean normal at that same locus. For the 28 loci identified as rod-mediated, RSL ranged from 4 to 44 dB (average, 14.4 dB). Rod sensitivity loss is less pronounced at greater eccentricities in the peripheral inferior field than closer to the scotomatous region, which is not only in the superior field but surrounds fixation. With LA perimetry (Fig. 2E), there is also an altitudinal scotoma; there are 30 extracentral loci with detectable LA function. A kinetic visual field (III-4e target) from the same eye of this patient shows an altitudinal scotoma; the central island and the inferior field island are separated by scotoma. Cone sensitivity loss (CSL), the difference between normal mean and patient results at each locus, ranged from no loss to 19 dB (average, 5.9 dB) for the 30 loci with detectable LA function. In summary, the technique in this patient shows detectable but mostly abnormal rod function in the inferior field and less affected cone function in this same region; there is a gradient of rod and cone function from more severe near the scotoma to less severe with eccentricity into the inferior field. 
Figure 2
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with autosomal dominant RP caused by a rhodopsin mutation (P15). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown below and to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown below and to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Figure 2
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with autosomal dominant RP caused by a rhodopsin mutation (P15). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown below and to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown below and to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Figure 3
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with X-linked RP caused by an RPGR mutation (P1). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Figure 3
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with X-linked RP caused by an RPGR mutation (P1). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Patient 1 is a 24-year-old man with X-linked RP (RPGR mutation). With DA perimetry and the blue stimulus, there were only eight extrafoveal loci detected, and these were in the nasal and temporal periphery (Fig. 3A). All of these peripheral loci were rod mediated (Fig. 3B); RSL ranged from 8 to 24 dB (average, 14.9 dB). With LA perimetry, there were 19 detectable extrafoveal loci, mainly in the nasal and temporal periphery. Kinetic perimetry with the III-4e target showed a central island and patches of nasal and temporal function with a complete annular midperipheral scotoma. Cone sensitivity loss ranged from 4 to 27 dB (average, 17.2 dB). Sensitivity at the foveal locus was also reduced. In summary, DA perimetry in this patient shows detectable but abnormal rod function only present in the peripheral field; cone function, also abnormal, is present in these peripheral regions as well as centrally. 
Comparing Results of the New Methodology With Those of Our Standard DA and LA Chromatic Perimetry in the Same IRD Patients
Of the 18 IRD patients examined with both DA perimetry techniques, a group of seven patients had >25 rod-mediated loci (average, 55 loci) as determined by the standard two-color method at the 70 extrafoveal loci of the RP test (Fig. 4A). The remaining 11 patients had ≤11 rod-mediated loci (average, 3.7 loci). The first group provided the opportunity to ask whether the rod-mediated loci with the standard method were all detected by the new method of DA perimetry. The answer was that most were, but some were not (Table). For example, the new DA method detected as rod mediated >95% of loci detected by the standard method in five patients (P18, P16, P15, P11, and P3). Two patients, P13 and P17, had detection of 79%. We tested the hypothesis that detected and undetected rod-mediated loci differed in degree of RSL (Fig. 4A). Mean RSL was different for detected and undetected loci in all patients (P < 0.05, t-test), with undetected loci having greater RSL. Mean differences between subsets ranged from 8.2 to 35 dB (average, 24 dB). 
Figure 4
 
Comparison of methodologies. (A) Boxplots of DA data from 7 patients, each with at least 25 loci determined to be rod mediated by standard two-color perimetry. The questions asked are how many of these loci are detected as rod mediated by the new DA method, and, when there are undetected loci, whether there is any difference in RSL in detected (white boxplots) versus undetected (gray boxplots) ones. Numbers within the graphic below boxplots are the detected versus undetected loci counts; the sum of the two numbers represents the total number of rod-mediated loci detected by the standard two-color perimetry. Band inside the box is the median; bottom and top of box are the first and third quartiles. Ends of whiskers represent the 10th and 90th percentiles. (B) Relationship of results of the two DA and the two LA methods. (C) Limits of agreement (95%) between the two DA methods and between the two LA methods. Mean differences of −3.2 dB for DA-500 nm minus DA-blue and −4.8 dB for LA-600 nm minus LA-white originating from differences of the definition of maximum (0 dB) stimulus have been removed from the plots shown in B and C.
Figure 4
 
Comparison of methodologies. (A) Boxplots of DA data from 7 patients, each with at least 25 loci determined to be rod mediated by standard two-color perimetry. The questions asked are how many of these loci are detected as rod mediated by the new DA method, and, when there are undetected loci, whether there is any difference in RSL in detected (white boxplots) versus undetected (gray boxplots) ones. Numbers within the graphic below boxplots are the detected versus undetected loci counts; the sum of the two numbers represents the total number of rod-mediated loci detected by the standard two-color perimetry. Band inside the box is the median; bottom and top of box are the first and third quartiles. Ends of whiskers represent the 10th and 90th percentiles. (B) Relationship of results of the two DA and the two LA methods. (C) Limits of agreement (95%) between the two DA methods and between the two LA methods. Mean differences of −3.2 dB for DA-500 nm minus DA-blue and −4.8 dB for LA-600 nm minus LA-white originating from differences of the definition of maximum (0 dB) stimulus have been removed from the plots shown in B and C.
We next analyzed the relationship between the results of the two methods (DA-500 nm versus DA-blue; LA-600 nm versus LA-white). Included in the analysis were 18 patients. All subjects were tested by both methods on the same day. For DA data, we only included locations with rod mediation (as determined by our standard method). Linear relationships between methods were evident for both DA and LA datasets (R2 = 0.90 and 0.87, respectively; P < 0.01, F-test; Fig. 4B). There were shifts of scale of −3.2 dB for DA and −4.8 dB for LA (mean difference, DA-500 nm minus DA-blue and LA-600 nm minus LA-white, respectively); this is not shown in the figure: 95% limits of agreement between methods were ±9.9 and ±6.8 dB for DA and LA data (Fig. 4C), respectively. As a reference, our previously reported intravisit variability value for DA static perimetry would correspond to limits of agreement for a difference of measurements of ±7.5 dB (2 SD variability limits, ±5.3 dB for a single measurement, DA-white, size V33). 
From DA and LA Perimetry Data Collection to Data Processing: A Plan
The many conventionally used tests in automated perimetry (e.g., for glaucoma) have instrument-based analyses. For the DA (and LA) testing we devised for IRD patients, the unmodified HFA is used solely to collect the data. The HFA 750i can be configured to export data from a test immediately after completion (Save and Transmit option) or at any time afterwards using the Transfer option within the File menu. In models having USB functionality, the exported files are written on a USB stick; for older models, they can be transferred using an RS232C to USB serial adapter. Either the memory stick or the serial adapter can be connected to a USB port in a standard PC (Windows, Mac, or Linux) or a mobile device (Android, IOS). Once in the laptop or mobile device, a downloadable program or app would read the files for rod and cone function analysis. In future multicenter clinical trials, the files can optionally be sent over the internet to a remote site for analysis or storage. In that case, the files would be coded for anonymization, digitally signed for integrity, and encrypted for privacy prior to data transfer (Fig. 5A). 
Figure 5
 
A plan for data processing for the DA and LA static perimetry results. (A) The HFA automated perimeter produces an XML-formatted file per test containing sensitivity values: one for DA and another for LA conditions. These files are transferred to a PC or mobile device for local analysis. The analysis programs would produce a single-page report for visualization of RSL and CSL, with contents similar to Figures 2 and 3 (not including F). Optionally, the system can be used to report to a reading center where original data can be stored and independently analyzed according to the center's own protocols. In this case, the local software can be used for anonymization and cryptographically secured data transfer. (B) Details of data processing. Measurements for locations with multiple samples are averaged. Losses are calculated by subtraction from mean normal. For RSL, only DA sensitivities with predicted mediation by rods are included. An additional longitudinal analysis can be constructed when historic data are available.
Figure 5
 
A plan for data processing for the DA and LA static perimetry results. (A) The HFA automated perimeter produces an XML-formatted file per test containing sensitivity values: one for DA and another for LA conditions. These files are transferred to a PC or mobile device for local analysis. The analysis programs would produce a single-page report for visualization of RSL and CSL, with contents similar to Figures 2 and 3 (not including F). Optionally, the system can be used to report to a reading center where original data can be stored and independently analyzed according to the center's own protocols. In this case, the local software can be used for anonymization and cryptographically secured data transfer. (B) Details of data processing. Measurements for locations with multiple samples are averaged. Losses are calculated by subtraction from mean normal. For RSL, only DA sensitivities with predicted mediation by rods are included. An additional longitudinal analysis can be constructed when historic data are available.
Rod and cone visual function analysis is shown as a flow diagram (Fig. 5B) and can run locally on a PC or mobile device or remotely using a web-accessible application. Data as received from the perimeter are parsed, and the following algorithm is applied for each location. First, a single sensitivity estimate is obtained for the location (by averaging, in the case of locations with multiple samples); second, sensitivity losses are obtained by subtracting the subject's sensitivity from corresponding mean normal sensitivity. For RSL, the measured sensitivity would first be determined to be from the rod system (far right branch of the diagram; Fig. 5B). This is necessary because DA perception at threshold may be mediated by cones in some cases. This assessment involves comparing the DA sensitivity against location-specific data for DA cone sensitivity (our DA results with the blue target at cone plateau). If the subject's DA sensitivity is higher than this value, perception at this location is assumed to be mediated by rods, and RSL is calculated by subtraction from mean DA normal sensitivity (for normal subjects, DA perception is mediated by rods throughout the retina). If the comparison fails, dark-adapted rod mediation cannot be ascertained, and no RSL value is calculated. A data visualization step summarizes the results (as exemplified in Figs. 2 and 3) including rod and cone sensitivities, sensitivity losses, and mediation. Multiple visit data can also be longitudinally compared with historical records separately for rods and cones. 
In summary, Figure 5 tries to reinform the readers that the commercially available automated perimeter, designed for glaucoma testing mainly, does not have analyses in the menu to cope with rod function sensitivity losses and does not fill the needs of the clinician-investigators who are seeking to understand the vision of IRD patients. Therefore, the specific data for IRD patients must be collected by the automated perimeter and then transferred to an external computer (e.g., PC). The latter device would have a suite of programs that can perform the step-by-step analysis shown in Figure 5B. Such programs would perform the needed analyses for many future purposes; not only determine rod and cone sensitivities across the visual field but allow visualization of the display as grayscale maps and answer key questions about intervisit variability and interocular differences and compare serial data of patients involved in either natural history studies or clinical trials of treatment. 
Discussion
Rod function is able to be measured with many methods (Supplementary Table S1). The method used in the clinic should fit the reason for making the measurement. The full-field electroretinogram (ERG) is a traditional diagnostic test that sums rod function across the entire retina.1 When the full-field ERG was used in clinical trials of RP patients at many different disease stages, the difficulties of obtaining recordable rod signals forced use of mainly cone ERGs as outcomes.34 A need to develop special techniques for recording small cone signals also occurred.35 There have also been attempts to use focal stimuli to elicit rod ERGs.36 Less commonly used assays with full-field stimulation are pupillometry,37,38 visual evoked cortical potentials (VECPs),39 and the full-field sensitivity test (FST).21,40 Visual evoked cortical potentials would be expected to elicit responses mainly from the central retina, considering the representation at the cortex.41 The FST, a psychophysical method, was designed specifically to detect the most sensitive retinal region (independent of location) and mainly in eyes with fixation abnormalities.21,40 
It has long been recognized that there is value in understanding regional variation of rod (and cone) function in retinal degenerations, but the current era of therapy has further emphasized the importance. There are therapies targeted to specific retinal regions42,43 and plans of treatment specifically for rod photoreceptors (e.g., Refs. 44 and 45). Not knowing the degree of rod functional impairment before introducing therapy and where in the retina this rod function is located (in diseases well known to have different patterns of visual loss32) precludes understanding whether a patient is a candidate for enrollment in specific clinical trials and, if a candidate, whether the therapeutic outcome is achieved. 
Traditionally, psychophysical measures of regional rod sensitivity have used relatively large targets such as with the Goldmann-Weekers adaptometer.31,46 Automated static perimetry, which is used routinely in most clinics as a method to detect regional variation in visual function, is most commonly performed in the LA state. The concept of DA static perimetry using chromatic stimuli to discriminate rod from cone function became a means to study and categorize RP and has remained a specialty test in a small number of clinics (e.g., Refs. 2, 12–16, and 47). Whether manual or automated perimeters are used, the goal for DA testing has been the same: to determine sensitivity levels of rod-mediated function at many loci in the visual field. What are the current alternatives to a computerized projection perimeter modified to deliver chromatic stimuli? There is an LED-based automated perimeter for dark-adapted two-color static perimetry (Medmont DAC) (Cideciyan AV, et al. IOVS 2016;57:ARVO E-Abstract 131), and it has the advantage of loci that cover most of the visual field. Disadvantages are current availability and also the issue of not being able to perform LA perimetry with the same instrument; the large number of loci in the instrument outweigh any disadvantage of being less flexible than a projection perimeter with custom testing options. 
Whereas most automated perimeters have algorithms to determine stability of fixation (e.g., blindspot monitoring) and also permit viewing of the patient's eye as a means to monitor fixation and cooperation, a method that allows performing perimetry while viewing the fundus is the ultimate way to localize the stimulus to the desired position on the retina. Fundus perimetry, as first named, has a decades-long history of being used in the DA state for measuring rod function in forms of retinal degeneration.48,49 Newer generations of fundus perimeters, some named microperimeters, have been devised and used mainly for measuring LA function in maculopathies (e.g., Ref. 50). More recently, advances have occurred such that there can also be testing in the DA state.51 The main disadvantage of this system is that it is limited to testing the central visual field. 
The advantages of the currently devised method of DA perimetry in an unmodified HFA are as follows: wide availability of the perimeter; familiarity to users in most eye clinics otherwise testing for glaucoma in the LA state with the size III target; flexibility of a projection perimeter to create custom tests; and the option to perform not only DA but also LA perimetry on the same instrument without time loss setting up the patient (and entering their identifying data) on another instrument. The disadvantages are as follows: the inability to determine rod mediation when rod sensitivities are reduced below normal cone sensitivities and to determine cone mediation in the DA state now that we are not using the two-color method; and, as in all automated perimetry, issues of free-viewing and fixation losses. 
The long-term goal is to develop not only a clinically feasible and accessible method for measuring rod (and cone) function in various parts of the visual field but also to coordinate the transfer of raw data to reading centers that will collect and analyze the data. A suite of computer algorithms for the analysis of two-color DA perimetry was developed and advanced over the last three decades for the data acquired by the modified HFA, but a revised analysis is now needed. This was initiated in the present report, but an extensive normal database is required. Also needed is more experience with use of the method and analysis of results in large cohorts of IRD patients who are age-matched to the normal subjects. A future and necessary goal will be to determine intervisit variability of the new method. The lack of such data represents a major limitation of the present study. Although not the focus of this work, the method may also be useful for studies of AMD.52,53 
Acknowledgments
Supported by grants from The Chatlos Foundation, Foundation Fighting Blindness, and Research to Prevent Blindness. 
Disclosure: D.B. McGuigan III, None; A.J. Roman, None; A.V. Cideciyan, None; R. Matsui, None; M.L. Gruzensky, None; R. Sheplock, None; S.G. Jacobson, None 
References
Berson EL. Retinitis pigmentosa and allied retinal diseases: electrophysiologic findings. Trans Sect Ophthalmol Am Acad Ophthalmol Otolaryngol. 1976; 81:OP659–OP666.
Massof RW, Finkelstein D. Rod sensitivity relative to cone sensitivity in retinitis pigmentosa. Invest Ophthalmol Vis Sci. 1979; 18: 263–272.
Rotenstreich Y, Fishman GA, Lindeman M, Alexander KR. The application of chromatic dark-adapted kinetic perimetry to retinal diseases. Ophthalmology. 2004; 111: 1222–1227.
Jacobson SG, Aleman TS, Cideciyan AV, et al. Identifying photoreceptors in blind eyes caused by RPE65 mutations: prerequisite for human gene therapy success. Proc Natl Acad Sci U S A. 2005; 102: 6177–6182.
Cideciyan AV, Swider M, Aleman TS, et al. Reduced-illuminance autofluorescence imaging in ABCA4-associated retinal degenerations. J Opt Soc Am A Opt Image Sci Vis. 2007; 24: 1457–1467.
Mitamura Y, Aizawa S, Baba T, et al. Correlation between retinal sensitivity and photoreceptor inner/outer segment junction in patients with retinitis pigmentosa. Br J Ophthalmol. 2009; 93: 126–127.
Mitamura Y, Mitamura-Aizawa S, Nagasawa T, et al. Diagnostic imaging in patients with retinitis pigmentosa. J Med Invest. 2012; 59: 1–11.
Cideciyan AV, Swider M, Jacobson SG. Autofluorescence imaging with near-infrared excitation: normalization by reflectance to reduce signal from choroidal fluorophores. Invest Ophthalmol Vis Sci. 2015; 56: 3393–3406.
Bramall AN, Wright AF, Jacobson SG, McInnes RR. The genomic, biochemical, and cellular responses of the retina in inherited photoreceptor degenerations and prospects for the treatment of these disorders. Annu Rev Neurosci. 2010; 33: 441–472.
Wright AF, Chakarova CF, Abd El-Aziz MM, Bhattacharya SS. Photoreceptor degeneration: genetic and mechanistic dissection of a complex trait. Nat Rev Genet. 2010; 11: 273–284.
Daiger SP, Sullivan LS, Bowne SJ. Genes and mutations causing retinitis pigmentosa. Clin Genet. 2013; 84: 132–141.
Zeavin BH, Wald G. Rod and cone vision in retinitis pigmentosa. Am J Ophthalmol. 1956; 42: 253–269.
Ernst W, Faulkner DJ, Hogg CR, et al. An automated static perimeter/adaptometer using light emitting diodes. Br J Ophthalmol. 1983; 67: 431–442.
Lyness AL, Ernst W, Quinlan MP, et al. A clinical, psychophysical, and electroretinographic survey of patients with autosomal dominant retinitis pigmentosa. Br J Ophthalmol. 1985; 69: 326–339.
Jacobson SG, Voigt WJ, Parel JM, et al. Automated light- and dark-adapted perimetry for evaluating retinitis pigmentosa. Ophthalmology. 1986; 93: 1604–1611.
Birch DG, Herman WK, deFaller JM, et al. The relationship between rod perimetric thresholds and full-field rod ERGs in retinitis pigmentosa. Invest Ophthalmol Vis Sci. 1987; 28: 954–965.
Jacobson SG, Apáthy PP, Parel JM. Rod and cone perimetry: computerized testing and analysis. In: Heckenlively JR Arden GB, eds. Principles and Practice of Clinical Electrophysiology of Vision. New York: Mosby-Year Book, Inc.; 1991: 475–482.
Cideciyan AV, Aleman TS, Boye SL, et al. Human gene therapy for RPE65 isomerase deficiency activates the retinoid cycle of vision but with slow rod kinetics. Proc Natl Acad Sci U S A. 2008; 105: 15112–15117.
Bainbridge JW, Smith AJ, Barker SS, et al. Effect of gene therapy on visual function in Leber's congenital amaurosis. N Engl J Med. 2008; 358: 2231–2239.
Dalkara D, Byrne LC, Klimczak RR, et al. In vivo-directed evolution of a new adeno-associated virus for therapeutic outer retinal gene delivery from the vitreous. Sci Transl Med. 2013; 5:189ra76.
Roman AJ, Schwartz SB, Aleman TS. et al Quantifying rod photoreceptor-mediated vision in retinal degenerations: dark-adapted thresholds as outcome measures. Exp Eye Res. 2005; 80: 259–272.
Johnson CA, Sample PA, Cioffi GA, et al. Structure and function evaluation (SAFE): I. criteria for glaucomatous visual field loss using standard automated perimetry (SAP) and short wavelength automated perimetry (SWAP). Am J Ophthalmol. 2002; 134: 177–185.
Bland JM, Altman DG. Agreement between methods of measurement with multiple observations per individual. J Biopharm Stat. 2007; 17: 571–582.
Myles PS, Cui J. Using the Bland-Altman method to measure agreement with repeated measures. Br J Anaesth. 2007; 99: 309–311.
Carstensen B, Simpson J, Gurrin LC. Statistical models for assessing agreement in method comparison studies with replicate measurements. Int J Biostat. 2008; 16: 1–26.
Pinheiro J, Bates D, DebRoy S, Sarkar D;,and the R Core Team. nlme: linear and nonlinear mixed effects models. R package version 3.1-124. Available at: http://CRAN.R-project.org/package=nlme. Accessed January 21, 2016.
Carstensen B, Gurrin L, Claus Ekstrom C, Figurski M. MethComp: functions for analysis of agreement in method comparison studies. R package version 1.22.2. Available at: https://CRAN.R-project.org/package=MethComp. Accessed January 21, 2016.
Berson EL, Rosner B, Sandberg MA, et al. Clinical trial of docosahexaenoic acid in patients with retinitis pigmentosa receiving vitamin A treatment. Arch Ophthalmol. 2004; 122: 1297–1305.
Berson EL, Rosner B, Sandberg MA. Clinical trial of lutein in patients with retinitis pigmentosa receiving vitamin A. Arch Ophthalmol. 2010; 128: 403 –4.
Birch DG, Weleber RG, Duncan JL, et al. Randomized trial of ciliary neurotrophic factor delivered by encapsulated cell intraocular implants for retinitis pigmentosa. Am J Ophthalmol. 2013; 156: 283–292.
Hoffman DR, Hughbanks-Wheaton DK, Spencer R, et al. Docosahexaenoic acid slows visual field progression in X-linked retinitis pigmentosa: ancillary outcomes of the DHAX trial. Invest Ophthalmol Vis Sci. 2015; 56: 6646–6653.
Grover S, Fishman GA, Brown J,Jr. Patterns of visual field progression in patients with retinitis pigmentosa. Ophthalmology. 1998; 105: 1069–1075.
Cideciyan AV, Hauswirth WW, Aleman TS, et al. Human RPE65 gene therapy for Leber congenital amaurosis: persistence of early visual improvements and safety at 1 year. Hum Gene Ther. 2009; 20: 999–1004.
Berson EL, Rosner B, Sandberg MA, et al. A randomized trial of vitamin A and vitamin E supplementation for retinitis pigmentosa. Arch Ophthalmol. 1993; 111: 761–772.
Andréasson SO, Sandberg MA, Berson EL. Narrow-band filtering for monitoring low-amplitude cone electroretinograms in retinitis pigmentosa. Am J Ophthalmol. 1988; 105: 500–503.
Sandberg MA, Pawlyk BS, Berson EL. Isolation of focal rod electroretinograms from the dark-adapted human eye. Invest Ophthalmol Vis Sci. 1996; 37: 930 –93.
Aleman TS, Jacobson SG, Chico JD, et al. Impairment of the transient pupillary light reflex in Rpe65(-/-) mice and humans with Leber congenital amaurosis. Invest Ophthalmol Vis Sci. 2004; 45: 1259–1271.
Kardon R, Anderson SC, Damarjian TG, Grace EM, Stone E, Kawasaki A. Chromatic pupillometry in patients with retinitis pigmentosa. Ophthalmology. 2011; 118: 376–381.
Jacobson SG, Knighton RW, Levene RM. Dark- and light-adapted visual evoked cortical potentials in retinitis pigmentosa. Doc Ophthalmol. 1985; 60: 189 –1.
Roman AJ, Cideciyan AV, Aleman TS, Jacobson SG. Full-field stimulus testing (FST) to quantify visual perception in severely blind candidates for treatment trials. Physiol Meas. 2007; 28: N51 –N5.
Creel DJ. Visually evoked potentials. In: Kolb H, Fernandez E, Nelson R, eds. Websion: The Organization of the Retina and Visual System. Salt Lake City, UT: University of Utah Health Sciences Center; 2013: 1–28.
Jacobson SG, Cideciyan AV, Ratnakaram R, et al. Gene therapy for Leber congenital amaurosis caused by RPE65 mutations: safety and efficacy in 15 children and adults followed up to 3 years. Arch Ophthalmol. 2012; 130: 9–24.
Jacobson SG, Cideciyan AV, Roman AJ, et al. Improvement and decline in vision with gene therapy in childhood blindness. N Engl J Med. 2015; 372: 1920–1926.
Lewin AS, Drenser KA, Hauswirth WW, et al. Ribozyme rescue of photoreceptor cells in a transgenic rat model of autosomal dominant retinitis pigmentosa. Nat Med. 1998; 4: 967–971.
Murray SF, Jazayeri A, Matthes MT, et al. Allele-specific inhibition of rhodopsin with an antisense oligonucleotide slows photoreceptor cell degeneration. Invest Ophthalmol Vis Sci. 2015; 56: 6362–6375.
Berson EL, Sandberg MA, Rosner B, Birch DG, Hanson AH. Natural course of retinitis pigmentosa over a three-year interval. Am J Ophthalmol. 1985; 99: 240–251.
Iannaccone A, Man D, Waseem N, et al. Retinitis pigmentosa associated with rhodopsin mutations: correlation between phenotypic variability and molecular effects. Vision Res. 2006; 46: 4556–4567.
Borruat FX, Jacobson SG. Advanced retinitis pigmentosa: quantifying visual function. Prog Clin Biol Res. 1989; 314: 3–17.
Jacobson SG, Cideciyan AV, Regunath G, et al. Night blindness in Sorsby's fundus dystrophy reversed by vitamin A. Nat Genet. 1995; 11: 27–32.
Cideciyan AV, Swider M, Aleman TS, et al. Macular function in macular degenerations: repeatability of microperimetry as a potential outcome measure for ABCA4-associated retinopathy trials. Invest Ophthalmol Vis Sci. 2012; 53: 841–852.
Crossland MD, Luong VA, Rubin GS, Fitzke FW. Retinal specific measurement of dark-adapted visual function: validation of a modified microperimeter. BMC Ophthalmol. 2011; 11: 1–6.
Owsley C, Jackson GR, Cideciyan AV, et al. Psychophysical evidence for rod vulnerability in age-related macular degeneration. Invest Ophthalmol Vis Sci. 2000; 41: 267–273.
Owsley C, McGwin G,Jr Clark ME, et al. Delayed rod-mediated dark adaptation Is a functional biomarker for incident early age-related macular degeneration. Ophthalmology. 2016; 123: 344–351.
Figure 1
 
Normal data for DA and LA RP tests using the unmodified HFA. (A) Dark-adapted blue normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD). (B) Dark-adapted blue at cone plateau. (Left) Mean normal. (Right) Dark adaptation functions indicating with brackets the concept of final DA thresholds (upper) and at cone plateau (lower). (C) Light-adapted white normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD).
Figure 1
 
Normal data for DA and LA RP tests using the unmodified HFA. (A) Dark-adapted blue normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD). (B) Dark-adapted blue at cone plateau. (Left) Mean normal. (Right) Dark adaptation functions indicating with brackets the concept of final DA thresholds (upper) and at cone plateau (lower). (C) Light-adapted white normal results. (Left) Mean normal. (Right) Lower limits of normal (mean − 2 SD).
Figure 2
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with autosomal dominant RP caused by a rhodopsin mutation (P15). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown below and to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown below and to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Figure 2
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with autosomal dominant RP caused by a rhodopsin mutation (P15). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown below and to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown below and to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Figure 3
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with X-linked RP caused by an RPGR mutation (P1). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Figure 3
 
Patient data using DA and LA RP tests. Results are from the right eye of a patient with X-linked RP caused by an RPGR mutation (P1). (A) Dark-adapted blue results. (B) Loci determined to be rod mediated, R. (C) Rod sensitivity loss at the identified rod-mediated loci. (D) Rod sensitivity loss map; grayscale shown to the right. (E) Light-adapted white results. (F) Kinetic visual field with III-4e target. (G) Cone sensitivity loss at the loci with detectable LA function. (H) Cone sensitivity loss map; grayscale shown to the right. N, nasal; T, temporal; I, inferior; S, superior visual field. x, physiologic blind spot locus. F, foveal-fixation locus.
Figure 4
 
Comparison of methodologies. (A) Boxplots of DA data from 7 patients, each with at least 25 loci determined to be rod mediated by standard two-color perimetry. The questions asked are how many of these loci are detected as rod mediated by the new DA method, and, when there are undetected loci, whether there is any difference in RSL in detected (white boxplots) versus undetected (gray boxplots) ones. Numbers within the graphic below boxplots are the detected versus undetected loci counts; the sum of the two numbers represents the total number of rod-mediated loci detected by the standard two-color perimetry. Band inside the box is the median; bottom and top of box are the first and third quartiles. Ends of whiskers represent the 10th and 90th percentiles. (B) Relationship of results of the two DA and the two LA methods. (C) Limits of agreement (95%) between the two DA methods and between the two LA methods. Mean differences of −3.2 dB for DA-500 nm minus DA-blue and −4.8 dB for LA-600 nm minus LA-white originating from differences of the definition of maximum (0 dB) stimulus have been removed from the plots shown in B and C.
Figure 4
 
Comparison of methodologies. (A) Boxplots of DA data from 7 patients, each with at least 25 loci determined to be rod mediated by standard two-color perimetry. The questions asked are how many of these loci are detected as rod mediated by the new DA method, and, when there are undetected loci, whether there is any difference in RSL in detected (white boxplots) versus undetected (gray boxplots) ones. Numbers within the graphic below boxplots are the detected versus undetected loci counts; the sum of the two numbers represents the total number of rod-mediated loci detected by the standard two-color perimetry. Band inside the box is the median; bottom and top of box are the first and third quartiles. Ends of whiskers represent the 10th and 90th percentiles. (B) Relationship of results of the two DA and the two LA methods. (C) Limits of agreement (95%) between the two DA methods and between the two LA methods. Mean differences of −3.2 dB for DA-500 nm minus DA-blue and −4.8 dB for LA-600 nm minus LA-white originating from differences of the definition of maximum (0 dB) stimulus have been removed from the plots shown in B and C.
Figure 5
 
A plan for data processing for the DA and LA static perimetry results. (A) The HFA automated perimeter produces an XML-formatted file per test containing sensitivity values: one for DA and another for LA conditions. These files are transferred to a PC or mobile device for local analysis. The analysis programs would produce a single-page report for visualization of RSL and CSL, with contents similar to Figures 2 and 3 (not including F). Optionally, the system can be used to report to a reading center where original data can be stored and independently analyzed according to the center's own protocols. In this case, the local software can be used for anonymization and cryptographically secured data transfer. (B) Details of data processing. Measurements for locations with multiple samples are averaged. Losses are calculated by subtraction from mean normal. For RSL, only DA sensitivities with predicted mediation by rods are included. An additional longitudinal analysis can be constructed when historic data are available.
Figure 5
 
A plan for data processing for the DA and LA static perimetry results. (A) The HFA automated perimeter produces an XML-formatted file per test containing sensitivity values: one for DA and another for LA conditions. These files are transferred to a PC or mobile device for local analysis. The analysis programs would produce a single-page report for visualization of RSL and CSL, with contents similar to Figures 2 and 3 (not including F). Optionally, the system can be used to report to a reading center where original data can be stored and independently analyzed according to the center's own protocols. In this case, the local software can be used for anonymization and cryptographically secured data transfer. (B) Details of data processing. Measurements for locations with multiple samples are averaged. Losses are calculated by subtraction from mean normal. For RSL, only DA sensitivities with predicted mediation by rods are included. An additional longitudinal analysis can be constructed when historic data are available.
Table
 
Clinical Characteristics of the IRD Patients
Table
 
Clinical Characteristics of the IRD Patients
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×