Reports

2 downloads 0 Views 3MB Size Report
30 Hz flashes (r= 0.62 for the entire sample) than they were for those elicited by .... 0.5 Hz flashes—thought to have eye-movement arti- facts because they had ...
Reports The Relationship Between Visual Field Size and Electroretinogram Amplitude in Retinitis Pigmentosa Michael A. Sandberg, Carol Weigel-DiFranco, Bernard Rosner, and Eliot L. Berson Purpose. To determine to what degree visual field size is correlated with electroretinogram (ERG) amplitude among patients with the common forms of retinitis pigmentosa (RP). Methods. Visual field equivalent diameter to the V4e white test light of the Goldmann perimeter was correlated with log ERG amplitude elicited by 0.5 Hz or 30 Hz full-field flashes of white light. Primary analyses were conducted on data from 583 patients with the common forms of RP. Subset analyses were performed on data from patients with ERG responses with different ranges of amplitude to assess to what extent the correlation depends on ERG amplitude, as well as on data from patients of a given genetic type to determine whether the correlation depends on die mode of transmission. Data from patients with the rhodopsin, Pro23His mutation (n = 38) or with the rhodopsin, Pro347Leu mutation (n = 24) were analyzed to determine the correlation between visual field size and ERG amplitude for patients with die same mutation.

Ijix previous studies'"6 have evaluated the correlation between visual field size and electroretinogram (ERG) amplitude for patients with retinitis pigmentosa (RP). Five of the six found that visual field size was significantly correlated with ERG amplitude (Table 1): the larger the field, the larger the ERG. Birch,3 Yagasaki,4 and lannacone et al6 found strong correlations—as high as 0.88 to 0.99. In the current study, we show that the correlation between visual field size and ERG amplitude, based on a sample of patients with RP larger than all previous samples combined, depends on various factors. Subset analyses were performed on patients with ERG responses within different ranges of amplitude to determine to what extent the correlation depends on the size of the ERG response. Analyses are presented on subsets of patients with different genetic types of RP and on patients with dominant disease and the rhodopsin, Pro23His or Pro347Leu mutation to determine whether the correlations are stronger when analyses are restricted to patients of a single genetic type or with the same mutation. PATIENTS AND METHODS. Patients. We first

analyzed the pretreatment data for 601 patients (age range, 18 to 49 years) with the common forms of RP who had participated in a clinical trial of nutritional Results. Visual field size was significantly correlated with supplements to slow the course of their disease (data ERG amplitude for every comparison (P =s 0.0003). set I). 7 Patients had (in at least one eye) a corrected Correlations generally were higher for ERGs elicited by visual acuity of 20/100 or better, a central visual field 30 Hz flashes (r= 0.62 for the entire sample) than they diameter of 8° or larger to the V4e white test light of were for those elicited by 0.5 Hzflashes(r = 0.53 for the the Goldmann perimeter, and an ERG amplitude of entire sample). They were lower for truncated ranges of 2.5 f£V or larger to 0.5 Hz full-field flashes of white ERG amplitude, higher for patients widi dominant or light and/or of 0.12 f£V or larger to 30 Hz full-field recessive disease than for patients with x-linked disease flashes of the same white light at a screening examinaor for patients of all genetic types combined, and strong tion. Pretreatment data were obtained on each eye for patients with the same rhodopsin mutation (reachseparately on two consecutive visits (screening and ing a value of 0.87). baseline) within 6 weeks and then averaged. Conclusions. Visual field size is significandy correlated We also analyzed data from a subset of patients with ERG amplitude for patients with RP. Correlation with dominantly inherited RP from a rhodopsin mutadepends on the range of ERG amplitudes, the inherition.8 There were no restrictions placed on eligibility tance type, and, particularly, on whether the analysis is for these patients. The subset (data set 2) included 39 confined to a single gene mutation. Invest Ophdialmol Vis Sci. 1996; 37:1693-1698. patients (age range, 23 to 73 years) with the rhodopsin, Pro23His mutation and 26 patients (age range, 8 to 63 years) with the rhodopsin, Pro347Leu mutation. These two groups of data set 2 were evaluated sepaFrom Ike Herman—Gtmd Uilxiratmy for the Study of Rr.lina.1 Degenerations, rately to determine the correlation between visual Harvard Medical School, Massachusetts Eye find Ear Infirmary, Hoston. field size and ERG amplitude among patients with Supported by grants from the National Eye. Institute (EY00I69 and EY02014) and Ity the h'imndat.um Fighting Blindness (Baltimore, Maryland). the same mutation. Before inclusion in this study, all froprietary interest category: N. patients signed informed consent forms. The study Submitted for publication October 17. 1995; revised March 12, 1996; accepted March 19. 1996. was approved by the Investigational Review Boards of Reprint requests: Michael A. Sandberg, Herman-Gund Udmratory, M0.68 (iV, the Pearson correlations were 0.52 and 0.52, respectively. Analysis by genetic type of data from patients in data set 1 is illustrated for the 30 Hz ERG condition in Figure 2. The Pearson correlation was 0.70 for patients with dominant disease and 0.72 for patients with autosomal recessive disease. For patients with x-linked disease, however, the Pearson correlation was 0.49, significantly smaller than that for the patients with dominant or recessive disease (P < 0.05 in each case). Strong relationships between visual field equivalent diameter and log ERG amplitude were found when analyses were performed on data from patients

0.1 »iV

1 .

140"

I

^v i

.

.

i

.

io jiv i

.

0.1

ioo

i

.

in data set 2 who had a particular rhodopsin mutation (Fig. 3). For example, the Pearson correlations for the 30 Hz ERG condition were 0.86 for patients with the rhodopsin, Pro23His mutation and patients with the rhodopsin, Pro347Leu mutation. In every instance except one, these correlations were significantly higher than those for all patients of data set 1 and for patients of data set 1 with dominant disease (P < 0.05 in each case); in the case that was the exception, Pearson correlations between visual field equivalent diameter and ERG amplitude to 30 Hz flashes for patients with the rhodopsin, Pro347Leu mutation (0.86) and patients with dominant disease from data set 1 (0.70) were not significantly different (P = 0.067).

10 nV I .

140'

0.1 uV

1 uV

110"

130-

• m

120-



110-

m

100" m



100"







"'A."

/



m

«

90-





^ r

80"

90-





'•

80-

'X

•"•-

^ ^ \ '

70"

'

'

• •

70-

60•

60-

' mi/

^

50-

\

'

\

50" 4030" 20-

rSX-' \?'- > •

•1.0

i *



40-







l -

a

*

30-

Dominant r=0.70, p