Foveal sensitivity changes in retinitis pigmentosa

Foveal sensitivity changes in retinitis pigmentosa Vivienne Greenstein, Donald C. Hood, and Ronald E. Carr

Loss in foveal sensitivity in retinitis pigmentosa (RP) has been attributed to a decrease in quantal catching

ability. Using a psychophysical technique (the probe-flash paradigm), we previously found that the results obtained for six RP patients under light adapted conditions were not consistent with a quantal catch hypothesis. To test further this hypothesis twelve RP patients were examined in dark adapted conditions. Probe thresholds were normal for five patients and increased for seven patients. The decreased quantal catching hypothesis was rejected for six of the seven patients.

1.

II.

Introduction

There is ample evidence from the psychophysical, densitometric, and electrophysiological literature that the cone system is affected in retinitis pigmentosa (RP). The loss in cone sensitivity has been attributed to a variety of factors. For example, since there is some evidence that RP leads to shortening of the photoreceptors and hence to a decrease in photopigment, it has been suggested that the concomitant decrease in quantal catching ability is responsible for the loss in visual sensitivity. Using a psychophysical paradigm, we found that the loss in foveal sensitivity could not be attributed

to a

decrease in quantum catching ability in functioning photoreceptors.1

Our results were surprising in view

of recent densitometric and psychophysical results favoring an explanation based on a decrease in quantal catching ability.2-4 Foveal densitometric studies have

demonstrated a reduced density of conepigment in RP patients. van Meel and van Norren4 found that the loss in sensitivity in their patients approximated the decrease in quantal absorption. In our experiment, the patients were tested in the light (a 24-cd/M2 steady background was used). As we do not fully understand

how RP affects the adaptational mechanisms, it seems prudent in view of the contradictory evidence to exam-

ine an additional group of RP patients in dark-adapted conditions. In this experiment the decreased quantal catching hypothesis is tested on RP patients in darkadapted conditions.

The authors are with Columbia University, Psychology Department, New York, New York 10027. Received 30 June 1986. 0003-6935/87/081385-00$02.00/0. © 1987 Optical Society of America.

Materials and Methods

A group of RP patients who had visual acuity of 20/20 and 20/30- (6/6 to 6/9-) clear media and suffi-

ciently large fields to ensure good foveal fixation throughout the test period were selected for the study. Table I shows the age, corrected Snellen acuity, and mode of inheritance of the twelve patients tested. All the patients had extinguished scotopic ERGs; photopic ERGs were either extinguished or the amplitudes were markedly reduced. Four subjects with no known abnormality of the visual system comprised the control group. All were in the 20-40-yr age group and had Snellen acuities in the tested eye of at least 20/20. A.

Apparatus

Light stimulation was provided by a projection system combined with an adapted Goldmann Weekers adaptometer. The final lens of the system projected circular images of the test and flash targets on a piece of frosted glass mounted in the sphere of the Goldmann Weekers adaptometer. Light intensity in the two channels was controlled by neutral density filters and by a two log unit neutral density wedge. All filters

were calibrated in the apparatus. The temporal characteristics of the stimuli were controlled by electronic shutters. B.

Stimuli

The spatial and temporal paradigms for the probeflash technique are shown in Fig. 1. The 500-ms flash was 10 in diameter, and the 10-ms probe light was 23

min of arc in diameter. The probe light was presented simultaneously with the onset of the flash. Foveal fixation was aided by four small red fixation lights forming a diamond pattern. The pattern subtended 20 and surrounded the stimuli. The flashes and adapting field were white (unfiltered tungsten light), and the probe was red (a cutoff filter was used which passed 10%of maximum at 600 nm and 1% at 598 nm). 15 April 1987 / Vol. 26, No. 8 / APPLIEDOPTICS

1385

Table I. Clinical Findings

2

3

Age

Visual acuity

Mode of

Patient

(yr)

tested eye

inheritance

1 2 3 4 5 6 7 8 9 10 11 12

11 22 37 42 35 50 25 42 14 30 36 40

20/20 20/30 20/3020/30 20/20 20/2020/20 20/30 20/20 20/2020/20 20/20

X-linked Simplex Simplex Simplex Simplex Simplex Simplex Autosomal dominanta Autosomal recessivea Simplex Simplexa Simplexa

a Thresholds within

3

,/P

O-s 0

T I

I 4

.

0 o

I

-1

I

I

Al3

NO FLASH

-3

-2

-1

0

1

LOG FLASH LUMINANCE (cd/n

Fig. 1.

2

2 )

Foveal probe-flash data for four normals obtained in dark

adapted conditions. The spatial and temporal paradigms are shown in the inset.

PATIENT 2

E -3

m

E

FLASH It en

I 2

normal limits.

PATIENT 1

B

IE

PATIENT3

i2

I. 2 B

*

a

z-

sa

0

a

3

a1

IO0 .

*

*

.

I1

t .

I

0

U U

*

0

8

I

.

L

O

,-,

-1

*

.

.

|

.

1.

c

-1

(3J '1

I o^ V)I

1

-

JS ' NO -3 FLASH

I

-2

-1

0

1

LOG FLASH LUMINANCE

Il

. 0

cd/n

2 2

)

-it 3 E

_1

U

O0

.

PATIENT4

FLASH

3 -2 -1 0 LOG FLASH LUMINANCE

PATIENT

5

(cd/n

12

.

I

's,

NO -3 -2 -1 0 12 FLASH LOG FLASHLUMINANCE (cd/n )

I )

3

E

- 2

2

3

PATIENT 6

- 2

a

a

. I TO

1 .

I

-

0

,J 0

*"

*

*

m

^1

*

_

11 -I

.

" s

,

NO -3 FLASH --

LU

-2 L...

, -1 - -..

I 0 -

LUMINANCE

o

AJ

1

2

. .

-1

3

No FLASH

(d/m')I

N3 E

-3

-2 1 0 1 2 ' Il LOG FLASH LUMINANCE (cd/ )

22

NO FLASH

-3 -2 -1 0 LOG FLASH LUMINANCE

1 (cd/rn)

2

PATIENT 7

0

D

(

I

I us

0

m

L

U

1

-I 0o

NoS FLASH

-3 -2 -1 0 LOG FLASH LUMINANCE

1 2 (cd/rn )

2

Fig. 2. Probe-flash data for seven RP patients. The squares in each panel represent the data obtained in dark-adapted conditions for each patient. The circles represent the median probe-flash data for the four normals shown in Fig. 1.

C. Procedure Throughout the experiment the subject's head was held in position by a chin rest and headrest. Subjects wore corrective lenses when necessary, the viewing distance of 30cm being taken into account. The natural pupil was used since mydriasis combined with an 1386

APPLIEDOPTICS / Vol. 26, No. 8 / 15 April 1987

artificial pupil did not significantly alter the data obtained on normals. The subject first dark adapted for 13 min; then increment thresholds were obtained for the probe in the presence of the larger longer duration flash. A method of limits procedure consisting of three to five up-and-down runs was used. The intensi-

patients can be seen as the large squares in Figs. 2(1),

A. DECREASED UANTALCATCH

(2), (3), (4), (5), (6) and (7). The small data points ()

3

represent the median probe thresholds for the four normals in Fig. 1. For each affected patient the threshold to the probe alone is increased compared with the normal. The mean increase in probe thresh-

E C- 2 -0 o

I

old is 0.57 log unit.

0

Ba

0I

I

Jo -4

-4

3:

3

A

-2

:

1

C

0

I

2

LOG FLASH LUMINANCE(d/.2I

IV. B. DECREASEDRESPONSIVENESS 3

E 2

'I

-2

I

I 0 X

B-i -I 0

-4

-3

-2

-1

O

I

2

LOG FLASH LUMINANCE cd/mr)

Fig. 3. Predicted probe-flash curves for two models of diseaserelated sensitivity loss. The continuous curves in (A) and (B) represent the probe-flash curves for the normal visual system in dark-

adapted conditions. The dashed curve in (A) represents the predicted change in the probe-flash curve for the decreased quantal catch model; all intensities are decreased by a multiplicative con-

stant. In (B) the dashed curve represents the predicted change for the decreased responsiveness model; the size of the response to all 5 flash intensities is scaled down by a multiplicative constant.

ty of the probe was varied in 0.1 log unit steps. The threshold for the probe alone (no-flash condition) was first obtained, then probe thresholds for eight flash intensities were measured. The flash intensities covered a range of -3.0 log units. Ill.

Results

Foveal probe-flash data for four normals are shown in Fig. 1. The log of probe threshold is plotted as a

function of log luminance of the flash. Probe threshold grows slowly at low-flash values, then rises with increasing steepness. Of the twelve patients tested five had probe thresholds which fell within the normal range. For seven RP patients (affected patients) probe thresholds were increased compared to the normal. The data for these

Patients

1, 2, and 3 show higher

probe thresholds for all flash intensities; patients 4, 5, and 6 show a similar pattern except at the highest flash intensity. A slightly different picture can be seen for patient 7 whose probe threshold falls within the normal range at the intermediate and higher flash intensities. Discussion

Are the data obtained in dark-adapted conditions consistent with an hypothesis of decreased quantum catching ability? Figure 3(A) shows the predicted probe-flash function for the hypothesis. The decreased quantal catching hypothesis predicts that the effective intensity of both the probe and flash are decreased by the same multiplicative constant. On a log-log plot the predicted probe-flash curve is shifted up and over by equal amounts compared to the normal [see Fig. 3(A)]. To compare the abnormal and normal probe-flash curves, the difference between probe thresholds can be measured as a function of flash intensity. The solid curve with small data points in Fig. 4 labeled decreased quantal catch presents the theoretical prediction in the form of a plot of the differences between the normal and expected abnormal probe thresholds (see figure caption for details). To derive these curves we calculated the difference between the normal data in Fig. 2 and the same data shifted vertically and horizontally by the patients' noflash threshold elevations. To verify our approach, we obtained probe-flash data from a normal subject with and without a 0.5 neutral density filter (see Fig. 5). Since the decreased quantal catching hypothesis is equivalent to placing a neutral density filter in front of a normal dark-adapted eye, probe-flash data obtained with a neutral density filter should resemble the prediction in Fig. 4(A) based on a sensitivity loss of 0.5 log

unit. They do; in Fig. 4(A), the open circles represent the differences between the normal probe thresholds obtained with and without the filter. The open circles in Fig. 4(A) fall very close to this curve. A comparison

of the patients' data to the theoretical prediction is shown in the other panels of Fig. 4. Only one of our seven affected patients (patient 7) has data which resemble the predictions of the decreased quantal catching hypothesis. This hypothesis does not describe the data for the other six patients.

Figs. 4. (A)-(H) Open circles represent the differences between the patient's log-probe thresholds and the median normal log-probe thresholds. The dashed horizontal lines represent the predicted difference based on the decreased responsiveness model; the predicted difference based on the decreased quantal catch model is represented by the interrupted curves (- *-). To derive the theoretical predictions we calculated the difference between the median data for normals in Fig. 2 and the same data shifted either vertically (the decreased responsiveness model) or vertically and horizontally (the decreased quantal catch model) by the patient's no-flash threshold elevations. For (A) a sensitivity loss of 0.5 log unit was assumed. Open circles in (A) represent the differences between probe threshold data obtained with and Figures on page 1388 without a 0.5-ND filter (see Fig. 5). For (B)-(H) the sensitivity loss ranged from 0.45 to 0.75 log unit. 15 April 1987 / Vol. 26, No. 8 / APPLIEDOPTICS

1387

w 1. 1a

The caption for Fig. 4 is on page 1387.

a -j

a -j zo 1.0 1-4

A.

zo

1.0

3I U)

W CI C.)W

DECREASED RESPONSIVENESS

w 0:

0.5

K*-o

WW

brn U-m U o 25 a:O

_~

~

~

DECREASED

-0.5 i_

0~~~~ .\--

lS~~

0

Cr

0

0.5

,.

I

NO

I

-3

-2

-1

-a

0

I

I

0

FLASH LOG FLASH LUMINANCE (cd/M

-j

0

W W

CD

0 -J -

0

0 0

b- j QUANTAL CATCH

K

~

B. PATIENT 1

2

)

0

D. -J ZO 1.0 r

C. PATIENT 2 1.0Co

"

L X

I-0.5-

00

b-n

I . NO FLASH

-0.5

1

C3

0

-2

-1

31

w

1

0

0

a - o.

- -0

0.51-

@R-----g

aN1

bWM o Lb 0

\.\

0

a.

(D

0

o-0. _E . .,

|

NO

I

is

-3

-J |

|

-2

I

I

-1

0

-LASH LOG FLASH LUMINANCE (cd/m

-0. 2

)

a

0: ls Ir 0.5

0

0

0 .

(cd/m 2 )

I

F. PATIENT 5

-J

zo 1.0r

1.0 0

NO -3 -2 -1 FLASH LOG FLASH LUMINANCE

1

E. PATIENT 4

-rn

I

PATIENT 3

Z 1,1Iw

U.m

I-J

I

0 (cd/m 2 )

LOG FLASH LUMINANCE

WI

0

I

I

-3

0 0

Z

o

0

0.5

t

crF 05 WW

-J

w o

0

U- 0

K0

a. 0 (D

0 0._

I

(4

NO

:

I

I

I

I

I

-3

-2

-1

0

1

FLASH LOG FLASH LUMINANCE (cd/m

a

I .

)

I

I

I

OK

O

H. PATIENT 7

0

Z 0:

I-. ,t 0.5 W-0.

-

0

I

-2 -1 FAHLOG FLASH LUMINANCE -

W n

E0. Uco

w - - --

c

~-

Lb-0

.

0 (d/rn

I 2

1

- --

_~~~~

CD

0 -j

0

1388

S

zo 1.0 r " XZ

w : .

-0.5

I

NO

0-J

G. PATIENT 6

r

M. 1

C')

2

- 0.5

I

I NO -3 -2 -1 0 FLASH LOG FLASH LUMINANCE (cd/rn ,

.

., ..

.

I

J

2

I

1 )

APPLIEDOPTICS / Vol. 26, No. 8 / 15 April 1987

-0.5

I

,.

NO FLASH

-3

-2

-1

LOG FLASH LUMINANCE

I

I

0

1

(cd/m 2 )

-

green. A stronger test of the decreased responsiveness hypothesis must await development of a paradigm

3

E

which will allow for maintenance of detection within a

.!2

single-cone pathway even at higher flash intensities. van Meel and van Norren4 concluded that the sensitivity loss in their RP patients was attributable to a loss in quantal catching ability. Their conclusions were based on a comparison in sensitivity loss measured psychophysically and pigment loss measured densitometrically. We do not understand the differences between their results and ours. It is likely that

S 1

B) I

A

4 0

A

0

A.

A

A

-1 0

l

l

NO FLASH

Fig. 5.

js

0 -1 -2 -3 LOG FLASH LUMINANCE

1 (cd/r2)

2

Foveal probe-flash data for a normal subject obtained with (A) and without (-) a 5-ND filter.

In an earlier study1 we concluded that the probeflash data obtained on six RP patients were consistent with an hypothesis of decreased responsiveness. That is, the data were consistent with a loss in sensitivity due to decreased responsiveness of retinal elements to light. We suggested that because of decreased metabolic activity or some other factor, the retinal cells responded to all light intensities with a fraction of the response produced in a nondiseased retina. The predicted probe-flash

function can be seen in Fig. 3(B).

To a first approximation, the decreased responsiveness hypothesis predicts a vertical shift in the dark adapted function.5 Figure 4 presents the prediction in the form of a plot of the differences between normal and diseased probe thresholds vs flash intensity. The dashed curves labeled decreased responsiveness indicate that the probe thresholds are increased by the same amount for all intensities. The data points for patients 1-5 fall closer to the decreased responsiveness curve than they do to the quantal catching curve. We cannot conclude, however, that they provide an adequate fit, especially at higher intensities. The deviations at higher intensities are due in part to the intrusion of different cone mechanisms. Subjects reported changes in the appearance of the probe at higher flash intensities; the color changed from red to white or pale

CHEMICAL PHYSICS

part of the sensitivity loss seen in our patients is due to

decreased quantal catching ability. In fact, the data for patient 6 can be fit by a combination of decreased quantal catching ability and decreased responsiveness. However, except for patient 7, we can conclude that a decrease in quantal catching ability accounts for at most a small percentage of the sensitivity loss in our patients. The authors also hold appointments with the Department of Ophthalmology of New York University.

This work was supported by National Eye Institute grant EY02115and by a grant from the RP Foundation and Allied Diseases to the NYU Retina Clinic. References 1. V. C. Greenstein, D. C. Hood, I. M. Siegel, and R. E. Carr, "Retinitis Pigmentosa: A Psychophysical Test of Explanations for Early Foveal Sensitivity Loss," Invest. Ophthalmol. Vis. Sci. 25, 118 (1984). 2. R. S. L. Young and G. A. Fishman, "Color Matches of Patients

with Retinitis Pigmentosa," Invest. Ophthalmol. Vis.Sci. 19,967 (1980). 3. R. S. L. Young and G. A. Fishmanj "Sensitivity Losses in a Long

Wavelength Sensitive Mechanism of Patients with Retinitis Pigmentosa," Vision Res. 22, 163 (1982). 4. G. J. van Meel and D. van Norren, "Foveal Densitometry in Retinitis Pigmentosa," Invest. Ophthalmol. Vis. Sci. 24, 1123 (1983). 5. D. C. Hood and V. C. Greenstein, "An Approach to Testing Alternative Hypotheses of Changes in Visual Sensitivity Due to Retinal Disease," Invest. Ophthalmol. Vis. Sci. 23, 96 (1983).

FIRST RESULTS FROM NBS BEAMLINE AT NSLS The first year of experiments using the NBS X-24A beamline at the Brookhaven National Synchrotron Light Source (NSLS) confirms that this facility will be an important national resource for x-ray spectroscopy. X-rays can be tuned through the energy range from 500 to 5000 eV to a bandwidth of 0.2 to 0.5 eV, two to five times better than competing sources, and focused to a 1- by 2-mm spot at the mple chamber.

Flux

at the sample

chamber

is 109 to 10

x-

rays/second. Current studies of argon, methyl chloride, various freons, and sulfur hexafluoride demonstrate that the instrumentation is sufficiently precise to separate xray spectral features due to multiple electron effects from those due to specific bond orbitals, long a stumbling block in such research. 15 April 1987 / Vol. 26, No. 8 / APPLIEDOPTICS

1389