Download SWS (blue) cone hypersensitivity in a newly identified retinal

Investigative Ophthalmology & Visual Science, Vol. 31, No. 5, May 1990
Copyright © Association for Research in Vision and Ophthalmology
SWS (Blue) Cone Hypersensitivity in a Newly
Identified Retinal Degeneration
Samuel G. Jacobson, Michael F. Marmor,* Colin M. Kemp, and Robert W. Knighron
Photoreceptor-mediated mechanisms were studied in patients with a recently identified retinopathy
typified by night blindness, cystoid maculopathy, and similar scotopic and photopic electroretinograms (ERGs). Dark-adapted spectral sensitivity functions were only partly explained as composites
of rod and cone curves shifted to lower sensitivities; there was unusually high sensitivity from 400-460
nm. A rod mechanism, reduced in sensitivity by at least 3 log units, was detectable with dark adaptometry. No measurable rhodopsin was found with fundus reflectometry. Light-adapted spectral
sensitivities were subnormal for wavelengths greater than 500 nm but supernormal from 420-460 nm.
On a yellow adapting field, the supernormal spectrum approximated that of the short-wavelengthsensitive (SWS) cone system. With spectral ERGs, two mechanisms were demonstrated. Dark- and
light-adapted ERGs to green, orange-yellow, and red stimuli had similar waveforms and coincident
intensity-response functions on a photopic intensity axis. ERGs to blue and blue-green stimuli were
similar, and intensity-response functions coincided on a SWS cone intensity axis. Patients varied in
the degree to which rod and midspectral cone function were decreased and SWS cone function was
increased. Invest Ophthalmol Vis Sci 31:827-838,1990
A retinal degenerative disease characterized by
night blindness, cystoid maculopathy, and atypical
electroretinograms (ERGs) with similar waveforms
under scotopic and photopic conditions has been
identified recently.1"3 Patients with this syndrome
have night blindness from early in life, varying degrees of visual acuity loss, and kinetic visual fields
that can be full or show defects centrally or in the
near midperiphery. Fundus pathology includes cystoid changes in the foveal region and small yellow
lesions and pigmentary disturbances near the vascular arcades. With clinical ERG techniques, there is:
no rod-mediated response to dimflashesof blue light
in the dark-adapted state; an abnormally reduced
cone flicker signal; mismatched responses to photopically balanced single flash stimuli with a larger signal to blue-green than to orange-yellow light; and
responses to single white light flashes that have simi-
lar waveforms whether elicited in the dark-adapted
state or on a white background light sufficient to suppress the rod system.3
Although the clinical features of this syndrome
have been well described,1"3 the pathophysiology of
the retinal dysfunction has not been fully elucidated.
In the current study, we used spectral sensitivity measurements, dark adaptometry, static perimetry, spectral ERGs, and fundus reflectometry to examine patients with different degrees of severity of this retinopathy. The results show very reduced but detectable
rod function, decreased sensitivity of middle-wavelength-sensitive (MWS) and long-wavelength-sensitive (LWS) cones, and surprisingly, a hypersensitivity
of the short-wavelength-sensitive (SWS) cone system.
Evidence is provided that the atypical ERG waveform in this retinopathy is mediated predominantly
by SWS cones.
Materials and Methods
From the Department of Ophthalmology, University of Miami
School of Medicine, Bascom Palmer Eye Institute, Miami, Florida,
and the *Department of Ophthalmology, Stanford University
School of Medicine, Stanford, California.
Supported in part by Public Health Service Research Grants
EY-05627 (SGJ) and EY-01678 (MFM), National Institutes of
Health, National Eye Institute (Bethesda, MD); an unrestricted
grant from Research to Prevent Blindness, Inc. (MFM); and the
National Retinitis Pigmentosa Foundation, Inc. (Baltimore, MD).
Submitted for publication: October 31, 1989; accepted December 19, 1989.
Reprint requests: Dr. S. G. Jacobson, Bascom Palmer Eye Institute, P.O. Box 016880, Miami, FL 33101.
Subjects
Three patients (patient 1, a 12-yr-old girl; patient 2,
a 29-yr-old man; and patient 3, a 40-yr-old woman)
from three different families (with no other known
affected members), and 26 normal subjects (16 female and 10 male; ages 13-59 yr) participated in this
study. Subjects gave their informed consent after full
explanation of the procedures was given. All patients
had small yellow lesions near the vascular arcades;
patients 2 and 3 had cystoid changes in the foveal
827
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / May 1990
828
region. Visual acuities in the right eye (the eye used
for psychophysical testing) of patients 1,2, and 3 were
20/25, 20/60, and 20/50 respectively, and color vision testing (Farnsworth D-15) was normal. Goldmann kinetic perimetry (with V-4e and I-4e test targets) in patient 1 was normal; in patient 2, there were
relative scotomas centrally and in the near midperiphery; and in patient 3, there was some superior
field limitation to the V-4e target and a nearly complete annular scotoma to I-4e in the midperiphery.
With conventional full-field clinical ERG techniques,4 all patients showed: no detectable signal to
dim blue flashes of light in the dark-adapted (>40
min) state; waveforms to singleflashesof bright white
light that were similar whether elicited in the darkadapted or light-adapted (34 cd/m 2 white background) states; an abnormally reduced cone flicker
(30 Hz) ERG; and unequal signals in response to
photopically balanced stimuli (amplitude to bluegreen light greater than that to orange-yellow).
Psychophysical Studies
A modified automated perimeter was used to perform static perimetry, dark adaptometry and measurements of spectral sensitivity in the right eye of all
three patients. Instrumentation, technique, analysis
methods, and normal results have been published
previously.5"7 Static thresholds were measured at 75
loci across the visual field in the dark-adapted state
using 500- and 650-nm stimuli and in the lightadapted state with a 440-nm stimulus (all target diameters, 103'; stimulus presentation time, 200 ms) on
Vol. 31
the white background light (10 cd/m2) of the automated perimeter.5 Dark-adapted sensitivity was also
measured with the 500-nm stimulus at 25 test loci in
a square area 20° on a side (5° grid) centered at 40°
nasally and 15° inferiorly in the visual field of patient
2. These results were compared with those of normal
subjects (n = 7, ages 19-59 yr) and then related to the
visual pigment measurements made in this retinal
region with imaging fundus reflectometry8'9 (see
below).
Dark adaptometry was performed with 420-nm
and 500-nm stimuli (target diameter, 103') at loci
with "mixed-mediation" (ie, rods detecting 500 nm
and cones detecting 650 nm) by two-color darkadapted perimetry.5610 Prebleach baselines were established after the patient had been dark-adapted for
at least 2 hr. After a full-field bleach, sensitivity was
measured for the two stimuli until the baseline levels
were attained.7
Spectral sensitivity was measured under three conditions: 1) in the dark-adapted (>2 hr) state; 2) in the
presence of the white background light of the automated perimeter (10 cd/m2); and 3) in the presence of
a yellow adapting field (two Lee filters, no. 179, over
fluorescent light sources situated at the edges of the
hemisphere; 123 cd/m2). Figure 1 shows the spectral
characteristics of the two background lights used in
this testing. The 15 monochromatic test stimuli (interference filters, 7.0-11.5 nm half bandwidth; Ealing
Electro-optics, Natick, MA) ranged from 400-680
nm in 20-nm steps. A United Detector Technology
40-X photometer and 1100B spectrometer were used
1.0
Fig. 1. Spectral characteristics of the three background lights used in this
study. Diamonds, the white
background light used for
light-adapted ERG recordings; plus signs, the white
background light of the automated perimeter; squares,
the yellow adapting field inserted into the perimeter.
440
600
WAVELENGTH (nm)
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
680
SWS (DLUE) CONE HYPERSENSITIVITY IN A RETINAL DISEASE / Jocobson er ol
No. 5
to determine the relative spectral output of the stimulus through the interference filters and of the background lights. Between three and six sensitivity measurements were made for each of the 15 test wavelengths, and the mean sensitivity was used for
plotting spectral sensitivity functions.
Spectral Electroretinography
Spectral ERGs were performed in patients 2 and 3
using a technique based upon published methods.""13 The instrumentation and methods we use to
record and analyze full-field ERGs have been described.714 ERGs were elicited with a series of intensities of blue, blue-green, green, orange-yellow, and
red light flashes (Wratten 98, 44, 61, 16, and 29
filters, respectively) in the dark-adapted state (>1.5
hr; stimulus frequency, 0.5 Hz) and in the lightadapted state (34 cd/m2 white background of the Nicolet GS-2000 Ganzfeld stimulator, following at least
15 min light adaptation; stimulus frequency, 1.0 Hz).
For each colored filter, there were six stimulus intensities, spanning 1 log unit at about 0.2-log-unit intervals; the maximum luminance of the unattenuated
white flash was 6.4 cd-s m~2. ERGs were also elicited
in the light-adapted state to blue and orange-yellow
light flashes flickering at 29, 35, and 39 Hz. The
spectral characteristics of the light-adapting background are shown in Figure 1.
Intensity axes on which colored filters were
weighted for various photoreceptive mechanisms
were calculated from:
700
X(X) F(X) W(X)
X=380
where
X(X) = the spectrum of the white flash;
F(X) = the spectral transmission of the colored filter;
W(X) = the weighting function of the photoreceptive
mechanism; and
the summation was in 10 nm steps.
For photopically weighting, W(X) = V(X), the CIE
photopic spectral luminous efficiency function15; for
scotopic weighting W(X) = V'(X), the CIE scotopic
luminous efficiency function15; and for SWS cone
weighting, W(X) = pi-3(X), Stiles's pi-3 field sensitivity function.15 The product X(X)F(X) was measured
in the ERG ganzfeld and was calculated from published data15 with similar results.
Imaging Fundus Reflectometry
Images obtained with a high-sensitivity television
(TV) imaging fundus reflectometer (IFR) were used
to estimate the levels of visual pigment present across
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
829
a circular area of retina of angular subtense 25° centered 36° temporally and 15° superiorly in the retina
of patient 2. Details of the IFR, which is based upon a
Zeiss 30° fundus camera to which a TV system has
been optically and mechanically coupled, have been
given elsewhere.8-9 Experimental methods and details
of data analysis also followed published procedures.8-9
The relative reflectances of fundal images acquired at
eight different wavelengths when the retina was fully
light-adapted were compared with those obtained
when the eye had been allowed to dark-adapt. The
density changes (the differences of the logarithms of
the reflectances for the two conditions) give an approximate measure of the level of visual pigment regenerated during dark adaptation.
Light adaptation was carried out with a white light
delivered by Maxwellian optics and covering an area
of retina concentric with but larger than the test area.
The light had an intensity of 6.0 log scotopic trolands.8-9 The duration of illumination used initially
was 1 min: in the normal eye such an exposure is
expected to remove more than 95% of the visual pigment.916 Recordings were made of the retinal images
obtained immediately after this exposure and at intervals for the following 30 min in darkness. In view
of the possibility that the subject possessed a visual
pigment of unusually low photosensitivity, the test
procedure was repeated after a further exposure to the
white adapting light, the duration of which was increased to 3 min.
Results
Psychophysical Studies
Figure 2 shows spectral sensitivity measurements
in the dark-adapted state at 36° in the inferior field
for normal subjects and the three patients. Although
the data from normals correspond well to the CIE
scotopic spectral sensitivity function, those of the patients do not. A CIE scotopic sensitivity function15
with its peak set about 3-3.5 log units below normal
tends to fit the wavelengths between 480 nm and 560
nm in all three patients. The patient data at longer
wavelengths can be fit by a portion of a published
peripheral cone function.17 For wavelengths less than
480 nm, however, there is higher sensitivity than expected from either of the functions. The sensitivities
for patient 3 at this inferior field locus are generally
lower than those for patients 1 and 2. Functions at
20° in the inferior field could be measured in patients
1 and 2 and were similar to those shown here, but
patient 3 had no detectable sensitivity at this locus.
To clarify whether the fully dark-adapted sensitivities measured at 420 nm and 500 nm were those of
rods or cones, dark adaptometry was performed with
Vol. 31
INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / May 1990
830
co
CD
~ZL
LU
CO
300
500
700
WAVELENGTH (nm)
and a yellow adapting field (Figs. 4C, D). With the
white background, spectral sensitivity in the normal
subjects has a peak at 440 nm and a second broad
maximum centered near 560 nm; sensitivity at 20°
(Fig. 4A) is generally higher than at 36° (Fig. 4B).
These curves are like those in published studies18"20
and are interpreted in terms of LWS, MWS, and SWS
cone systems. The patient data differ from those of
the normal subjects by having higher sensitivity at the
shorter wavelengths and lower sensitivity for wavelengths greater than 520 nm at both test loci. There
are also differences between the data of the patients.
At 20° inferiorly, there was no measurable sensitivity
for patient 3; sensitivities at wavelengths less than 480
nm in patient 2, although slightly higher than normal, are lower than those of patient 1. At 36° inferiorly, the results for patients 1 and 2 are similar, but
patient 3 shows sensitivities lower than those of the
other patients. It is also of note that the shape of the
patients' light-adapted spectral sensitivity functions
differs from that measured in the dark-adapted state
(Fig. 2). In the light-adapted function, sensitivity at
500 nm is decreased relative to lower and higher
wavelengths; this not only provides further evidence
of a rod mechanism but also implies that the rods are
suppressed by the background light.
Fig. 2. Spectral sensitivity measurements, dark-adapted, in: normal subjects, (n = 5; ages 19-29 yr; area between solid lines represents range of data); patient 1, PI; patient 2, P2; and patient 3, P3.
Squares represent patient data. Solid curves in P1-P3 are CIE scotopic sensitivity curves15 shifted vertically and equated to the patient data at 500 nm. Dashed curves represent a peripheral cone
function17 equated to the patient data at 600 nm.
o102030-
these two colors (Fig. 3). In a representative normal
subject, sensitivities for both 420 nm and 500 nm
increase rapidly in the first few minutes and then
reach a plateau, the typical pattern for the "cone
branch" of the dark adaptation function. After about
12 min, a further increase in sensitivity occurs and
continues until about 40 min after the bleach, representing the "rod branch" of the function. In the data
of patient 2, sensitivity to the 420-nm stimulus does
not change after the first few minutes, but sensitivity
to the 500-nm stimulus slowly increases and reaches
prebleach baseline levels after about 1 hr. A similar
pattern of results was obtained in patient 3, but sensitivity to 500 nm did not reach the baseline for about
90 min. In all three patients, the difference between
sensitivity for 500 nm at the cone plateau and at final
dark-adapted sensitivity was about 0.5-0.7 log units.
Apparently, the dark-adapted sensitivity was rod-mediated at 500 nm but cone-mediated at 420 nm.
Figure 4 shows spectral sensitivity measurements
at two eccentricities in the inferior field performed in
the presence of a white background light (Figs. 4A, B)
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
4050CO
TD
60-
N
P2
70
40
20
40
60
O-i
2
LU
in
102030405060-
P3
70
20
40
60
ao
100
TIME (min)
Fig. 3. Dark adaptometry with 420 nm (squares) and 500 nm
(plus signs) stimuli in: a 29-yr-old normal subject (N) at 20° inferior field; patient 2 (P2), also at 20° inferior; and patient 3 (P3) at
36° inferior and 12° temporal.
T=>
831
SWS (BLUE) CONE HYPERSENSITIVITY IN A RETINAL DISEASE / Jocobson er ol
No. 5
6-
440
520
600
680
440
520
600
WAVELENGTH (nm)
Fig. 4. Spectral sensitivity measurements performed with a white background light (A, B) and with a yellow adaptingfield(C, D) in normal
subjects (area between solid lines represents range of data; for A and B, n = 5, ages 19-29 yr; for C and D, n = 3, ages 24-29 yr), patient 1
(squares, dotted lines), patient 2 (plus signs, dashes of equal length), and patient 3 (diamonds, dashes of unequal length). (A, C) Data from 20°
inferior field. (B, D) Data from 36° inferior field.
With the yellow adapting field, spectral sensitivity
in the normal subjects peaks at 440 nm; between 500
nm and 600 nm, there is measurable, although very
reduced, sensitivity. The normal function at 20° inferiorly (Fig. 4C) has generally higher sensitivity than
that at 36° (Fig. 4D). These findings resemble those
from earlier studies and represent selective adaptation of LWS and MWS cones.21"23 At 20°, sensitivity
at 440 nm is higher than the mean normal by about
0.8 log units in patient 1 and by about 0.3 log units in
patient 2, but sensitivity beyond 520 nm is not detectable; patient 3 had no measurable sensitivities at this
locus. At 36°, patients 1 and 2 have sensitivity at 440
nm nearly 1 log unit higher than the mean of the
normals, whereas patient 3 is about 0.5 log units
higher; again, sensitivity is unmeasurable beyond 520
nm. The spectral sensitivity functions with the yellow
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
adapting field in the patients are very similar to published functions for the SWS cone system.2122'24
Figure 5 shows gray-scale maps of dark-adapted
sensitivity to 500 nm and to 650 nm and lightadapted sensitivity to 440 nm across the visualfieldof
the patients and normal subjects. The normal mean
map for the 500-nm stimulus (Fig. 5, left column, N),
representing rod sensitivity,51017 is relatively flat, decreasing slightly at increasing eccentricities. In the 3
patients, sensitivity differences between 500 and 650
nm indicated that there were no rod-mediated loci,
but only mixed-mediated or cone-mediated loci.56
All of the test loci had abnormally reduced sensitivity
to 500 nm (ie, >2 SD below the mean at each locus),
and the maps indicate that the loss was relatively diffuse and severe across the visual field. Patients 2 and
3 show more profound losses than the younger pa-
832
INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / May 1990
500 nm, dark-adapted
650 nm, dark-adapted
Vol. 31
4 40 nm, light-adapted
m
P3
Fig. 5. Gray-scale maps of dark-adapted sensitivity to 500 nm (left) and 650 nm (center) and light-adapted sensitivity to 440 nm (right), in
the right eyes of patient 1 (PI), patient 2 (P2), and patient 3 (P3). The normal results (N) represent the mean data from 16 subjects (ages 13-59
yr) tested with the 500-nm stimulus, dark-adapted; 4 subjects (ages 22-41 yr) tested with the 650-nm stimulus during the cone plateau of dark
adaptation; and 6 subjects (ages 19-29 yr) with 440 nm, light-adapted. The gray scales used to represent the results all have 16 levels of gray
(key at lower right). Maximum sensitivity, shown as white, is 54 dB for 500 nm, 34 dB for 650 nm, and 20 dB for 440 nm; black indicates no
detection of the stimulus.
tient 1 and also show scotomatous areas (ie, "0 dB"
sensitivity, illustrated as black). Mean extrafoveal rod
sensitivity loss for patient 1 is 34.0 dB (SD = 3.4); for
patient 2, 41.7 dB (SD = 4.2); and for patient 3, 45.8
dB (SD = 4.6). In the patient data, many of the loci in
the central field were cone-mediated,56 and therefore
the mean rod sensitivity losses may be underestimates.
The normal mean map for 650 nm (Fig. 5, central
column) was measured at the cone plateau (4-8 min
after the bleach), thereby representing longer wavelength cone sensitivity across the visual field. The
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
decline in sensitivity from the central to the far peripheral field is consistent with published data.17 In
the patients, since there were no rod-mediated loci,
dark-adapted sensitivity to 650 nm represents longer
wavelength cone sensitivity. Patient 1 shows a relatively mild cone sensitivity loss (25% of loci were >2
SD below the mean), but patient 2 has more loss (63%
of loci abnormally reduced), especially superiorly and
in the temporal periphery. Patient 3 shows the most
sensitivity loss (89% of loci abnormally reduced) with
scotomatous regions superiorly, temporally, and nasally.
No. 5
SWS (DLUE) CONE HYPERSENSITIVITY IN A RETINAL DISEASE / Jocobson er ol
The regional retinal variation of the increased sensitivity at shorter wavelengths in the patients was investigated by performing perimetry with a 440-nm
stimulus on a white background light (Fig. 5, right
column). In normal subjects, sensitivity decreases
with increasing eccentricity, a pattern which is consistent with previously published work.1923 Patients 1
and 2 do not show the relatively steep decrease in
sensitivity with eccentricity that is present in the normal, but instead show a more flat distribution of sensitivity that decreases only in the far periphery. For
patient 1, 25% of loci were within 2 SD of the mean
normal, and 75% of the loci had sensitivities that were
at least 2 SD higher than normal. For patient 2, the
foveal locus was subnormal; 33% of the loci were
normal; and the remaining 65% were higher than
normal. Patient 3 shows sensitivity losses in the same
regions of the visual field (mainly superior, temporal,
and nasal) as were evident with 650 nm, darkadapted. Sixteen percent of the loci were >2 SD
below the normal mean; 77% were normal; and 7%
were higher than normal. Although there is an apparent increase in severity of dysfunction with age in
these three patients, a direct relationship between age
and clinical measures of severity was not evident in a
larger series of patients.3
Spectral Electroretinography
ERGs were recorded in the light- and dark-adapted
states using different colored stimuli (Fig. 6), and the
results were analyzed to determine if the signals from
the patients represented one or more than one photoreceptive mechanism (Fig. 7). Figure 6 (upper)
shows the light-adapted ERGs in a representative
normal subject and in patients 2 and 3. In the normal
subject, the largest response is obtained with the orange-yellow stimulus, with the next largest from
blue-green and green, and then red, and finally blue
light. In contrast, patients 2 and 3 show the largest
amplitude responses to the blue and the blue-green
stimuli; there are smaller responses to orange-yellow
and green and a barely measurable response to red. It
is also of note that the patients' ERGs to the blue and
blue-green stimuli differ dramatically from those of
the normal by having a broader waveform and slower
implicit time. Although less dramatic, the patients'
responses to the other colors are also not normal in
waveform appearance.
Figure 6 (lower) illustrates that in the normal subject the dark-adapted ERGs to blue, blue-green,
green, and yellow stimuli are relatively equal in amplitude; the response to red light is smaller and shows
the typical early cone and later rod components.4 The
ERGs from patients 2 and 3 are lower in amplitude
than those of the normal; responses to blue and blue-
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
800
green light flashes have the largest amplitudes, with
the other colors showing lower amplitude signals.
The patients' dark-adapted ERGs resemble their
light-adapted recordings in waveform shape; the amplitudes in the dark tend to be larger, especially for
the middle- and long-wavelength (green, orange-yellow, and red) stimuli.
The principle of univariance25 predicts that if responses to two different spectral stimuli arise from a
single photoreceptive mechanism, then: 1) the intensity-response functions for the two stimuli should coincide when plotted on an intensity axis that weights
the stimuli for the mechanism under consideration;
and 2) responses of equal amplitude to the two different stimuli should have similar waveforms. By use of
this principle, the spectral ERGs of the patients were
divided into two groups. Figure 7 shows results from
patient 2; the results for patient 3 were similar. In the
light-adapted (Fig. 7A) and dark-adapted (Fig. 7B)
states, the middle- and long-wavelength stimuli produced responses with similar waveforms that had
coincident intensity-response functions on a photopically (MWS + LWS cone) weighted intensity axis.
The short-wavelength stimuli (blue and blue-green)
produced responses with similar waveforms and
coincident intensity-response functions on an intensity axis weighted for SWS cones (Fig. 7C, D). Although the same grouping applied in both light- and
dark-adapted conditions, dark-adapted ERGs may
have had a small rod component, because responses
to blue-green are slightly more sensitive relative to
blue than in the light-adapted ERGs (compare Fig.
7D to Fig. 7C). Figure 7C also shows the lightadapted response function to blue light in the normal
subject (probably representing SWS and midspectral
cone systems13). The much larger signals in the patient again suggest hypersensitivity of the SWS cone
system.
Reported differences in critical flicker fusion frequency of the SWS cone system compared with that
of MWS and LWS cones26"29 prompted us to compare ERGs elicited with blue and orange-yellow lights
in the light-adapted state at 29, 35, and 39 Hz in
patient 2. Stimulus intensities that elicited lightadapted responses of similar amplitude to the two
colors at a frequency of 1 Hz were chosen to elicit
responses at the higher frequencies. At 29 Hz, there
was a definite response to both colors, but at 35 and
39 Hz there was only a response to the orange-yellow
stimulus. Equal amplitude responses to 35 Hz flicker
could be obtained with the aforementioned orangeyellow stimulus and a higher-intensity blue stimulus
(by 0.4 log units); the two responses, however, differed in timing. With this higher intensity blue stimulus, there were still no detectable responses at 39 Hz.
834
Vol. 31
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / May 1990
B
B-G
O-Y
R
Light-adapted
N
P2
P3
Dark-adapted
N
P2
P3
Fig. 6. ERGs elicited with different intensities of blue (B), blue-green (B-G), green (G), orange-yellow (O-Y), and red (R) light flashes in a
19-yr-old normal subject (N), in patient 2 (P2), and in patient 3 (P3). Upper three sets of traces: ERGs elicited on a white background light
from the right eyes of N and P2 and the left eye of P3. Lower three sets of traces: ERGs elicited in the dark-adapted state from the right eyes of
N, P2, and P3. The three stimulus intensities in relative log units are " 0 " or maximum (top), —0.4 (middle), and — 1.0 (bottom). Calibration is
at lower right: vertically, 100, 50, and 25 nV for N, P2, and P3 respectively in light-adapted recordings; and 200, 50, and 25 n~V for N, P2, and
P3 respectively in dark-adapted recordings. Horizontally, 20 msec for all traces. Stimulus onset occurs 20 msec after trace onset.
Imaging Fundus Reflectometry
No measurable visual pigment was detected within
the measurement area during the IFR testing of patient 2: all double density changes observed at 7 wavelengths ranging from 460-600 nm9 were within 1 SD
of zero. This was the case both when the retinal illuminance of the bleaching-adaptation exposure was
7.8 log scotopic troland-sec, and when it was raised to
8.3 log scotopic troland-sec. The mean sensitivity loss
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
with the 500 nm stimulus within the area tested with
IFR was about 30 dB. The lack of measurable visual
pigment and this degree of rod sensitivity loss are
consistent with the relationship expected if decreased
quantal absorption were the sole cause of the rod
dysfunction.9
Discussion
In contrast to other retinal diseases which can show
decreased SWS function,3031 the retinopathy de-
SWS (DLUE) CONE HYPERSEN5ITMTY IN A RETINAL DISEASE / Jocobson er ol
No. 5
200-
100
-3
835
-2
-2
-1
-1
SWS units
Photopic units
LOG
RELATIVE
INTENSITY
Fig. 7. Intensity-response functions for spectral ERGs of patient 2. (A, C) Light-adapted (LA). (B, D) Dark-adapted (DA). Responses fell
into two groups with different underlying photoreceptive mechanisms. (A, B) B-wave amplitudes of responses to longer wavelength stimuli
(open triangles, green; circles, orange-yellow; open squares, red), plotted against photopically weighted intensity. (C, D) B-wave amplitudes of
the responses to short wavelength stimuli (filled triangles, blue; asterisks, blue-green) plotted against intensity weighted for SWS cones. (C) also
shows the intensity-response function of a light-adapted normal subject (filled squares) to blue stimuli. In each panel are waveforms from
responses indicated by the arrows. Stimulus onset, which is 20 msec after trace onset, is shown as a small arrow on the baseline preceding the
waveforms.
scribed herein appears to be unique in manifesting
increased function of the SWS visual pathway. Evidence of SWS cone hypersensitivity was provided by
results of both psychophysical and electrophysiologic
studies. With measures of dark- and light-adapted
spectral sensitivity, the effect was present; the finding
was confirmed by isolating the SWS mechanism with
chromatic adaptation. These psychophysical tests
also indicated that the midspectral cone system was
abnormally low in sensitivity, and that rod function
was so severely impaired as to be barely measurable.
Electrophysiologic correlates to these findings were
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
provided by applying the principle of univariance to
spectral ERG data. The patients' dark- and lightadapted ERGs were shown to represent mainly a supernormal SWS cone mechanism and a subnormal
midspectral cone mechanism; what may be a small
rod system component was detectable in the darkadapted recordings.
Whether the retinal locus for the increased SWS
cone function and that for the decreased midspectral
cone function are at the receptor level, or whether
they involve postreceptoral mechanisms,32"33 is not
known. The IFR results, however, serve to localize
836
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / May 1990
the rod dysfunction to the receptor. Given the low
sensitivity of the rod system, at least two possible
outcomes might have been expected from the IFR
testing. If this disorder were like that in congenital
stationary night blindness,34 normal levels of rhodopsin would have been present. The finding of no
measurable rhodopsin suggests that the operating
range is set by the reduced ability of the rods to absorb (or to photolyse after absorption of) incident
quanta. A prolonged time course of rod adaptation
such as was demonstrated in these patients has been
reported also in vitamin A deficiency35 and in retinitis pigmentosa.36 A serum vitamin A level in patient 3
was normal.
The findings in the current study provide a means
to interpret the symptoms of patients with this retinopathy and to interpret the results of their visual
function tests performed with clinical techniques.3
The 3-4 log units of rod sensitivity loss measured
perimetrically (and the unmeasurable rhodopsin
levels) are in accord with the complaints of night
blindness and the nondetectable rod ERG b-wave to
dim blue flashes in the dark-adapted state. The midspectral cone sensitivity losses across much or all of
the field are consistent with the abnormally reduced
cone flicker ERG. A full Goldmann kinetic visual
field in some of the patients (eg, patient 1), despite
severely impaired rod function and abnormal midspectral cone sensitivity, is understandable, given that
the SWS system is functioning at supernormal levels
across most of the field. When field defects are present by Goldmann perimetry (eg, patients 2 and 3),
they are in regions of dysfunction of all the photoreceptor systems: the midspectral cone and rod systems
are subnormal and the SWS system is subnormal or
normal rather than supernormal in sensitivity. Apparently, enough MWS and LWS function remains
to perform the Farnsworth D-15 test correctly. Undoubtedly, more refined tests of color vision would
reveal differences from normal. The decreased visual
acuity could simply be the result of a "dropping-out"
of the MWS and LWS cones, but further study is
required to clarify the basis of the central retinal dysfunction and its relation to the cystoid macular
changes.
The atypical ERG waveform reported previously
as "rod-like" under dark- and light-adapted conditions1'2 must represent predominantly a supernormal
SWS cone system ERG with a subnormal midspectral cone component. Although much larger than
SWS cone ERGs isolated in normal subjects with
chromatic adaptation or silent substitution, the ERG
with the SWS cone mechanism in the patients of the
current study has the same broad waveform shape,
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
Vol. 31
implicit time in the range of 45-65 msec, and a relatively low critical flicker fusion frequency. Unlike the
published normal SWS cone ERGs,37"39 there is an
a-wave at higher stimulus intensities and a relatively
steep b-wave amplitude versus log intensity function.
Study of this waveform under conditions known to
elicit specific phenomena from SWS cone ERGs (eg,
transient tritanopia, off-response behavior) would be
of interest for future work.3840
The combination of retinal function abnormalities
in this disorder is unlike that determined with similar
techniques in either typical retinitis pigmentosa9 or
the cone-rod dystrophies.7'41 The results of retinal
function tests in three atypical retinopathies,213'42
however, do share features with those described
herein, although interpretations of the results differ.
Gouras et al13 performed spectral ERGs in patients
with a "unique cone-rod degeneration" and explained the results (some of which resembled those in
the current study) as an altered electric responsiveness of rods and cones and a defect in rod saturation.
Unpublished psychophysical data were said to show
insensitive rods but a normal SWS cone system. The
lack of psychophysical evidence for a supernormal
SWS mechanism may have resulted from testing a
retinal region with substantial dysfunction. Keunen
et al42 reported two siblings with a "congenital stationary night blindness" that showed no measurable
rhodopsin and identical photopic and scotopic ERG
parameters. There also were abnormal results with
foveal cone densitometry but no fundus abnormalities; these findings raise the possibility that central
cone dysfunction might precede the cystoid maculopathy. In a 14-yr-old patient described as having a
"rod-cone dystrophy" by Fishman and Peachey,2
clinicalfindingswere like those described for this retinopathy; ERGs evoked with blue and orange stimuli
light-adapted were interpreted as originating from the
rod system and a defect in rod adaptation was proposed.
To test the hypothesis that insensitive and poorly
saturating rods were the basis of the ERG responses,213 we plotted all of our light-adapted responses on a scotopically weighted axis (Fig. 8). Intensity-response functions for most of the colors did
not coincide, but the overlap at certain intensities of
the functions to blue-green and orange-yellow stimuli
may explain why Fishman and Peachey,2 who used
only these two colored stimuli, concluded that their
recordings were from the rod system. Figure 8 (waveform inset) shows, however, that amplitude-matched
responses to blue-green and orange-yellow stimuli are
quite different in timing, and therefore arose from
different mechanisms. Clearly, a rod mechanism does
No. 5
SWS (BLUE) CONE HYPERSENSITMTY IN A RETINAL DISEASE / Jocobson er ol
200
-2
-1
0
Scotopic units
LOG
RELATIVE
INTENSITY
Fig. 8. Intensity-response functions for light-adapted (LA) spectral ERGs of patient 2, plotted on a scotopically weighted axis.
Open triangles, green; circles, orange-yellow; open squares, red;
filled triangles, blue; asterisks, blue-green. Waveforms shown are of
equal amplitude responses to blue-green (top trace) and orangeyellow (bottom trace) stimuli.
not govern the light-adapted ERG responses in our
patients.
The three patients in this study differed in degree of
severity of the retinal dysfunction. The fundal
changes near the vascular arcades and the cystoid
maculopathy 3 certainly suggest that the disease process has a degenerative component, at least in these
retinal regions. Longitudinal studies of retinal function are needed to determine if the differences among
patients represent slow progression of a retinal degeneration, variable expression, or both.
Key words: nightblindness, electroretinography, retinal degeneration, short-wavelength-sensitive cones, spectral sensitivity
Acknowledgments
The authors thank A. J. Roman, M. I. Roman, and P. P.
Apathy for help with data collection and analyses; J-M.
Parel, W. Lee, and D. Denham for modifications to the
perimeter to permit spectral sensitivity measurements; Mrs.
M. G. Robey for coordinating this research; and R. S. L.
Young, PhD, for his critical comments on this manuscript.
References
1. Marmor MF: Large rod-like photopic signals in a possible new
form of congenital night blindness. Doc Ophthalmol 71:265,
1989.
2. Fishman GA and Peachey NS: Rod-cone dystrophy associated
with a rod system electroretinogram obtained under photopic
conditions. Ophthalmology 96:913, 1989.
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
837
3. Marmor MF, Jacobson SG, Foerster MH, Kellner U, and Weleber RG: Diagnostic clinicalfindingsof a new syndrome with
night blindness, maculopathy and enhanced blue cone sensitivity (submitted).
4. Gouras P: Electroretinography: Some basic principles. Invest
Ophthalmol 9:557, 1970.
5. Jacobson SG, Voigt W, Parel J-M. Apathy PP, Nghiem-Phu L,
Myers SW, and Patella VM: Automated light- and darkadapted perimetry for evaluating retinitis pigmentosa. Ophthalmology 93:1604, 1986.
6. Apathy PP, Jacobson SG, Nghiem-Phu L, Knighton RW, and
Parel JM: Computer-aided analysis in automated darkadapted perimetry. In Seventh International Visual Field
Symposium, Greve EL and Heijl A, editors. Dordrecht, the
Netherlands, Martinus NijhofT Publishers/Dr W Junk Publishers, 1987, pp 279-284.
7. Yagasaki K and Jacobson SG: Cone-rod dystrophy: Phenotypic diversity by retinal function testing. Arch Ophthalmol
107:701, 1989.
8. Faulkner DJ and Kemp CM: Human rhodopsin measurement
using a TV-based imaging fundus reflectometer. Vision Res
24:221, 1984.
9. Kemp CM, Jacobson SG, and Faulkner DJ: Two types of
visual dysfunction in autosomal dominant retinitis pigmentosa. Invest Ophthalmol Vis Sci 29:1235, 1988.
10. Massof RW and Finkelstein D: Rod sensitivity relative to cone
sensitivity in retinitis pigmentosa. Invest Ophthalmol Vis Sci
18:263, 1979.
11. Gouras P, Eggers HM, and MacKay CJ: Cone dystrophy, nyctalopia and supernormal rod responses: A new retinal degeneration. Arch Ophthalmol 101:718, 1983.
12. Gouras P, MacKay CJ, Ivert L, Mini RN, Neuwirth J, and
Eggers H: Computer-assisted spectral electroretinography in
vitrectomy patients. Ophthalmology 92:83, 1985.
13. Gouras P, MacKay C, Evers H, and Eggers H: Computer assisted spectral electroretinography (CASE): A tool for examining hereditary retinal degenerations. In Retinal degeneration:
Experimental and clinical studies, Hollyfield J and La Vail M,
editors. New York: Alan R. Liss, Inc, 1985, pp. 115-130.
14. Jacobson SG, Yagasaki K, Feuer WJ, and Roman AJ: Interocular asymmetry of visual function in heterozygotes of X-linked
retinitis pigmentosa. Exp Eye Res 48:679, 1989.
15. Wyszecki G and Stiles WS: Color science: Concepts and
methods, quantitative data and formulae, 2nd ed. New York:
John Wiley & Sons, 1982.
16. Ripps H, Brin KP, and Weale RA: Rhodopsin and visual
threshold in retinitis pigmentosa. Invest Ophthalmol Vis Sci
17:735, 1978.
17. Stabell B and Stabell U: Absolute spectral sensitivity at different eccentricities. J Opt Soc Am 71:836, 1981.
18. Wooten BR, Fuld K, and Spillman L: Photopic spectral sensitivity of the peripheral retina. J Opt Soc Am 65:334, 1975.
19. Verriest G and Uvijls A: Central and peripheral increment
thresholds for white and spectral lights on a white background
in different kinds of congenitally defective colour vision. Atti
Fond Giorgio Ronchi 32:213, 1977.
20. Kuyk TK: Spectral sensitivity of the peripheral retina to large
and small stimuli. Vision Res 22:1293, 1982.
21. Stiles WS: Color vision: The approach through incrementthreshold sensitivity. Proc Natl Acad Sci USA 45:100, 1959.
22. Wald G: The receptors of human color vision. Science
145:1007, 1964.
23. Wooten BR and Wald G: Color vision mechanisms in the
peripheral retinas of normal and dichromatic observers. J Gen
Physiol 61:125, 1973.
838
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / May 1990
24. Baylor DA, Nunn BJ, and Schnapf JL: Spectral sensitivity of
cones of the monkey Macaca fascicularis. J Physiol 390:145,
1987.
25. Naka KJ and Rushton WAH: S-potentials from colour units in
the retina offish (Cyprinidae). J Physiol 185:536, 1966.
26. Brindley GS, DuCroz JJ, and Rushton WAH: The flicker fusion frequency of the blue-sensitive mechanism of colour vision. J Physiol 183:497, 1966.
27. Green DG: Sinusoidal flicker characteristics of the color-sensitive mechanisms of the eye. Vision Res 9:591, 1969.
28. Yoshida T and Kobayashi Y: Blue cone system in human
spectral ERG. Acta Soc Ophthalmol Jpn 83:2055, 1979.
29. Yagasaki K and Miyake Y: Clinical application of the responses of the blue cone system studied with electroretinogram
(ERG): analysis of recording conditions and responses in normal subjects. Acta Soc Ophthalmol Jpn 87:98, 1987.
30. Young RSL: Early-stage abnormality of foveal pi mechanisms
in a patient with retinitis pigmentosa. J Opt Soc Am 72:1021,
1982.
31. Greenstein VC, Hood DC, Ritch R, Steinberger D, and Carr
RE: S (blue) cone pathway vulnerability in retinitis pigmentosa, diabetes and glaucoma. Invest Ophthalmol Vis Sci
30:1732, 1989.
32. Pugh EN and Mollon JD: A theory of the pi-1 and pi-3 color
mechanisms of Stiles. Vision Res 19:293, 1979.
33. Zrenner E: Electrophysiological characteristics of the blue sen-
Downloaded From: http://iovs.arvojournals.org/ on 08/03/2017
34.
35.
36.
37.
38.
39.
40.
41.
42.
Vol. 31
sitive mechanism: Test of a model of cone interaction under
physiological and pathological conditions. Doc Ophthalmol
Proc Series 33:103, 1982.
Ripps H: Nightblindness revisited: From man to molecules.
Invest Ophthalmol Vis Sci 23:588, 1982.
Kemp CM, Jacobson SG, Faulkner DJ, and Walt RW: Visual
function and rhodopsin levels in humans with vitamin A deficiency. Exp Eye Res 46:185, 1988.
Alexander KR and Fishman GA: Prolonged rod dark adaptation in retinitis pigmentosa. Br J Ophthalmol 68:561, 1984.
Norren DV and Padmos P: Human and macaque blue cones
studied with electroretinography. Vision Res 13:1241, 1973.
Evers HU and Gouras P: Three cone mechanisms in the primate electroretinogram: Two with and one without off-center
bipolar responses. Vision Res 26:245, 1986.
Sawusch M, Pokorny J, and Smith VC: Clinical electroretinography for short wavelength sensitive cones. Invest Ophthalmol
Vis Sci 28:966, 1987.
Valeton JM and van Norren D: Transient tritanopia at the
level of the ERG b-wave. Vision Res 19:689, 1979.
Ripps H, Noble KG, and Greenstein VC, et al: Progressive
cone dystrophy. Ophthalmology 94:1401, 1987.
Keunen JEE, van Meel GJ, and van Norren D: Rod densitometry in night blindness: A review and two puzzling cases. Doc
Ophthalmol 68:375, 1988.