Frequency Doubling Perimetry in Resolved Optic Neuritis

Abstract

purpose. To study the visual field with frequency doubling technology (FDT) in patients with recovered optic neuritis and to detect loss of magnocellular projecting cells (M cells) in the extrafovea.

methods. Fourteen patients who had undergone one attack of optic neuritis and recovered normal vision (1.0 or better) and critical fusion frequency were examined with conventional Humphrey automated perimetry central 30-2 and FDT c-20 threshold tests. After 1 year, 12 patients were reexamined with central 30-2 and FDT c-20 tests. The visual fields examined by both perimeters were divided into three zones. The mean sensitivity in each zone in involved eyes, uninvolved eyes, and involved eyes after 1 year was compared with that in healthy eyes.

results. Conventional automated perimetry showed depression toward the fovea. However, FDT demonstrated general depression, especially midperipheral deficits. After 1 year, the midperipheral deficits with frequency doubling perimetry (FDP) improved, as did central depression, as observed with central 30-2 tests.

conclusions. FDT was developed to detect early glaucomatous damage, which was thought to be caused by a loss of M cells. Our study suggested that patients with resolved optic neuritis also had a loss of M-cell function in the extrafoveal area, as observed by field damage and its recovery.

Patients with optic neuritis have shown abnormalities of critical flicker frequency, ¹ contrast sensitivity, color vision, and central visual field ^{2 3} after their visual acuity improved. Except visual field, these tests reflected foveal function. Wall reported that parvocellular projecting retinal ganglion cell (P cell) function was more affected in the foveal area than the magnocellular projecting cell (M cell) function in resolved optic neuritis. ² P-cell function reflected color vision and contrast sensitivity at high spatial frequency and low temporal frequency. However, Jacobson and Olson ¹ showed impaired critical flicker frequency, which reflected M-cell function, in the fovea in recovered optic neuritis. In the extrafoveal region, sensitivity at some temporal frequency was checked at maximal eccentricity of 10° in optic neuritis. ⁴ Recently, frequency doubling technology (FDT) has been developed to detect glaucomatous damage. ^{5 6} This perimetry, using frequency doubling illusion, reflected a subclass of M-cell (M_Y) function even at an eccentricity of 20°. To our knowledge, no reports have described M-cell function in optic neuritis at the extrafoveal area. We studied M-cell function at a range of 20° in patients with recovered optic neuritis with the use of FDT and compared the results with conventional automated perimetry. 

Materials and Methods

Fourteen patients (6 men and 8 women) who had one attack of decreased vision (range, 0–0.5) and recovered their vision and critical fusion frequency to normal ranges were studied. Their ages ranged from 15 to 55 years old (mean, 35.4 years). Ten patients had an attack of decreased vision in one eye, and 4 patients had binocular loss of vision. Only one eye in these 4 patients was randomly selected. Fourteen healthy volunteers (7 men and 7 women) were included in this study. They ranged in age from 24 to 50 years (mean, 36.7 years). All had 1.0 or better visual acuity and 35 Hz or more of critical fusion frequency, as measured by CFF Test Apparatus II (Matsumoto Medical Instruments, Osaka, Japan). The stimulation subtended 2° of arc. Critical fusion frequency was determined when the subjects perceived the beginning of flicker. Three descending trials were averaged. The conventional visual field was performed with Humphrey field analyzer 750 (Zeiss-Humphrey, Dublin, CA). The central visual field was measured by threshold central 30-2 (76 test points in central 30° field) program. Frequency doubling perimetry was measured by FDT (Humphrey-Welch Allyn, Dublin, CA). The program was threshold c-20 (17 test areas in 20° field). One test area was 10° of square or arc (the very center). Each area had black and white 0.25 cycle per degree sinusoidal grating at 25 Hz flickering. The contrast between black and white varied automatically with the response of the subjects. Contrast sensitivity was calculated from log contrast and expressed as dB. Both visual fields were measured twice after recovery, and the second result was adopted. Twelve patients were reexamined with both perimeters after 1 year. Results with more than 30% of false positive, false negative, or fixation loss were excluded. The visual field was divided into three zones, which were composed of 1 test area (R1; the very center), 4 test areas (R2; inside 10° field), and 12 test areas (R3; inside 20° field) of FDT (Fig. 1A ). The Humphrey 30-2 visual field was also divided into three zones corresponding to the three respective FDT zones (Fig. 1B) . Age, critical fusion frequency, and mean sensitivity in each zone in involved eyes, uninvolved eyes, and involved eyes after 1 year was compared with those in healthy eyes using one-way ANOVA and post hoc tests (Bonferroni-Dunn method). Statistical significance level was 0.0167 for critical fusion frequency and 0.0083 for age and mean sensitivity. 

The research was conducted in accordance with institutional guidelines and with the tenets of the World Medical Association Declaration of Helsinki. We obtained written informed consent from each subject. 

Results

All data are shown in Table 1 . 

Critical fusion frequency was 44.1 ± 1.1 (mean ± SE) Hz in eyes that had a previous attack of optic neuritis, 46.3 ± 0.85 Hz in unaffected eyes, and 48.9 ± 1.2 Hz in healthy eyes. Critical fusion frequency in involved eyes was significantly (P = 0.0021) decreased, compared with that in healthy eyes. 

Fourteen visual fields by Humphrey 30-2 were estimated as 8 with normal findings and 6 with localized defects (2 peripheral rim, 1 paracentral, 1 central scotoma, and 2 enlarged blind spots), according to criteria of the Optic Neuritis Treatment Trials. ⁷ None had dense central scotomas. 

Mean sensitivity in R1 and R2 with Humphrey 30-2 in involved eyes was significantly lower than that in each respective zone in healthy eyes (Fig. 2A ). Mean sensitivity in R3 in involved eyes was not significantly decreased. After 1 year, mean sensitivity in R1 and R2 with Humphrey 30-2 in 12 involved eyes was also significantly lower than that in healthy eyes. Statistical P level (0.0002) in mean sensitivity in R1 decreased to 0.0025 after 1 year. 

Mean sensitivity in R2 and R3 with FDT in involved eyes was significantly lower than that in each respective zone in healthy eyes (Fig. 2B) . After 1 year, mean sensitivity in each zone with FDT in 12 involved eyes was not significantly lower than that in healthy eyes. 

No significant abnormalities in mean sensitivity of each zone examined with both perimeters in uninvolved eyes were found, compared with that in healthy eyes. 

Discussion

Conventional Humphrey automated perimetry showed general depression in patients with resolved optic neuritis, especially within central depression in a 5° field. The findings corresponded with those reported by Wall. ² FDT demonstrated general depression in resolved optic neuritis, especially depression in the extrafoveal area. The area outside 10° of eccentricity showed an abnormality with the Humphrey field analyzer. ² This perimeter could not separate the function of P and M cells. M-cell function was related to high temporal frequency. Sensitivity up to 10° field at temporal frequency was abnormal at up to 2.5° at 23 Hz in optic neuritis. ⁴ The area outside 10° at high temporal frequency had not previously been examined in optic neuritis; however, FDT could cover up to 20° at high temporal frequency. Glaucomatous visual field defects were specifically detected by FDT. ^{5 6} Early glaucomatous damage was correlated with a loss of large retinal ganglion cells (M cells) ⁸ and nonlinear units (M_Y cells). ⁹ Patients in the present study had normal intraocular pressure throughout the course and no glaucomatous cupping after resolution. Visual field defects in optic neuritis were thought to include losses of both P and M cells. Both P and M cells were found in the extrafoveal area and in the fovea. ¹⁰ Central depression detected by conventional automated perimetry led to a loss of P cells, ¹ and deficits in the midperiphery detected by FDT led to a loss of M_Y cells, because of dissociation of field damage pattern and its long-term recovery observed between both perimeters. Central (R1) depression was detected by conventional automated perimetry, but mean sensitivity in R3 was not significantly decreased. On the other hand, midperipheral deficits (R2 and R3) were found by FDT, but the central area (R1) was near normal. Midperipheral deficits by FDT tended to recover after 1 year, whereas the field in R3 observed by conventional automated perimetry remained unchanged. FDT takes approximately 5 minutes per eye to perform and could be useful as a follow-up study of the midperipheral visual field of patients with optic neuritis after recovery. 

One third of the uninvolved eyes showed abnormal visual fields in the recovery stage. ⁷ The visual field in uninvolved eyes by conventional automated perimetry and FDT showed slight general depression but was not significantly different in any zone. 

 Supported in part by a grant from the Ministry of Education, Japan, Grant 10357015.

 Submitted for publication November 29, 1999; revised February 23, 2000; accepted March 22, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

(A) Three zones (R1, R2, and R3) in FDT threshold c-20 are shown as right eye. R1 is the central circle at central 5°. R2 is 4 areas between 5 and 10°. R3 is 12 squares between 10 and 20°. (B) Three zones in Humphrey threshold central 30-2 corresponding with those in FDT c-20 are shown.

Table 1.

 

Summary of Statistical Data

Parameter	Healthy Eyes (n = 14)	Involved Eyes (n = 14)	Uninvolved Eyes (n = 10)	Involved after 1 Year (n = 12)	Involved vs Healthy (P)	1 Year vs Healthy (P)
Age (y)	36.7 ± 2.4	35.4 ± 3.5	40.5 ± 2.5	36.1 ± 3.9	0.7643	0.8934
CFF (Hz)	48.9 ± 1.2	43.9 ± 1.1	46.3 ± 0.85	NA	0.0021	NA
HFA R1 (dB)	32.7 ± 0.28	29.9 ± 0.50	31.6 ± 0.40	30.4 ± 0.80	0.0002	0.0025
HFA R2 (dB)	31.2 ± 0.35	29.3 ± 0.50	30.6 ± 0.28	29.2 ± 0.79	0.0065	0.0052
HFA R3 (dB)	29.8 ± 0.36	27.3 ± 0.90	29.0 ± 0.48	27.4 ± 1.30	0.0610	0.0426
FDT R1 (dB)	32.5 ± 0.76	28.7 ± 1.30	31.0 ± 1.4	28.8 ± 0.70	0.0109	0.0150
FDT R2 (dB)	31.5 ± 0.68	28.1 ± 1.40	30.2 ± 1.0	28.8 ± 1.00	0.0077	0.0385
FDT R3 (dB)	29.9 ± 0.69	26.1 ± 1.50	28.3 ± 0.81	27.5 ± 0.96	0.0013	0.0419

 Values are means ± SE. CFF, critical fusion frequency; HFA, Humphrey field analyzer; FDT, frequency doubling technology. A P value with an underline indicates statistical significance. Statistical analysis was done with one-way ANOVA and post hoc test.

Figure 2.

 

(A) The mean sensitivity in three zones by Humphrey 30-2 corresponding to three respective zones by FDT c-20. The mean sensitivity in R1 and R2 in involved eyes was significantly lower than that in R1 (P = 0.0002) and R2 (P = 0.0065), respectively, in healthy eyes. The mean sensitivity in R1 and R2 zones in involved eyes after 1 year was also significantly decreased. *Statistically significant decreased mean sensitivity. Bars, SEs. (B) The mean sensitivity in three zones by FDT c-20. The mean sensitivity in R2 and R3 in involved eyes was significantly lower than that in R2 (P = 0.0077) and R3 (P = 0.0013), respectively, in healthy eyes. The mean sensitivity in three zones after 1 year was not significantly decreased. *Statistically significant decreased mean sensitivity. Bars, SEs.