Long-term visual prognoses in patients with retinitis pigmentosa: the Ludwig von Sallmann lecture

Abstract

Retinitis pigmentosa can be followed over almost its entire course with narrow bandpassed, computer averaged cone electroretinograms (ERGs). The long-term rate of decline of these responses can be described by an exponential function. A cone ERG actuarial table based on 1039 patients and 6553 visits is presented to show the estimated number of years for an average patient with a given 30-Hz cone ERG amplitude to decline to 0.05 microV (i.e. virtual blindness). The table is based on a projected rate of loss of 10% of remaining cone ERG amplitude per year for those not on treatment and 8.3% per year for those on treatment with vitamin A palmitate 15,000 IU/day. The table can be used to provide an estimate of the average long-term visual prognosis from a single visit; more precise estimates for a specific patient require several additional visits over 2- to 3-year intervals. Evidence is presented to support the idea that patients with a projected cone amplitude of 3.5 microV or greater at age 40 (about 25% of our patient population with typical retinitis pigmentosa) would be expected, on average, to retain some useful vision for their entire lives without treatment. Knowledge of the amount of remaining cone function in the ERG often reduces patient anxiety and helps patients plan for their future.

Figure 1

Full-field (Ganzfeld) system in use for ERG testing. Stroboscope light (Grass PS2), enclosed in a case and attached to the top of top of diffusing sphere, illuminates the inner white surface of this dome, 40cm in diameter, providing a full-field stimulus. (From Rabin and Berson, 1974).

Figure 2

Full-field ERG responses from a normal subject and from four patients with retinitis pigmentosa. Stimulus onset is vertical hatched lines for columns 1 and 2 and vertical shock artifacts for column 3. Rod b-wave implicit times in column 1 and cone implicit times in column 3 are designated with arrows. Calibration symbol (lower right corner) signifies 50 msec horizontally and 100 μV vertically for all tracings. Under these test conditions, normal amplitudes are ≥ 100 μV (left column), ≥350 μV (middle column), and ≥50 μV (right column). Normal rod implicit time is ≤108 msec and normal cone implicit time is ≤32 msec. (From Berson, 1976.)

Figure 3

Full-field ERG responses from a normal subject and four patients with self-limited sector retinitis pigmentosa or stationary forms of night blindness. Horizontal arrows (column 1) designate range of normal rod b-wave implicit times, and vertical bar defining this range (mean ± 2 SD) has been extended through responses of patients with sector retinitis pigmentosa. Responses (middle column) from patient with Oguchi’s disease are interrupted by reflex blinking, so latter part cannot be illustrated. Cone implicit times in column 3 are designated with arrows. (From Berson, 1976.)

Figure 4

Full-field ERG responses recorded in 1967 and 1977 from a normal subject and from four affected members of a family with a dominant form of retinitis pigmentosa. Pedigree number and age at time of testing are indicated for each patient. One to three responses to the same stimulus are represented. Calibration symbol for 1967 responses designates 50 msec horizontally for columns 1 and 2 and 25 msec for column 3 and 50 μV vertically for column 1 and 100 μV vertically for columns 2 and 3. Calibration symbol for 1977 responses designates 60 msec horizontally and 100 μV vertically for all tracings. (From Berson and Simonoff, 1979.)

Figure 5

Full-field 30-Hz cone ERGs from a normal subject and four patients with retinitis pigmentosa tested at an 11–15-year interval. Stimulus onset, vertical markers; calibration symbol (left column, lower right) designates 100 μV vertically for the normal subject and top three patients and 40 μV vertically for the bottom patient and 50 msec horizontally for all traces; calibration (right column, lower right) designates 2 μV vertically for the dominant, X-linked, and isolate patients and 0.3 μV for the recessive patient and 20 msec horizontally for all traces. B-wave implicit times are designated with arrows. (From Andreasson, et al., 1988.)

Figure 6

Narrow bandpassed, computer-averaged full-field cone ERGs in response to 30 Hz flashes (n=256 sweeps with 6 responses per sweep) from a representative patient with retinitis pigmentosa at baseline and 6 years after baseline. Three consecutive averages are superimposed. The broken vertical line indicates the onset of the train of flashes. (From Berson, et al., 1993)

Figure 7

Frequency of positive responses on questionnaire as a function of 30-Hz cone electroretinogram (ERG) amplitude for 586 patients with retinitis pigmentosa. Groups were defined based on quintile of ERG amplitude: <0.34 μV (n=132); 0.34 to 0.67 μV (n=125); 0.67 to 1.40 μV (n=117); 1.40 to 4.53 μV (n=109); and >4.53 μV (n=103). (From Berson, et al., 1993).

Figure 8

Regression of cone ERG amplitude on age among 1039 patients with retinitis pigmentosa in the database of the Berman-Gund Laboratory. Patients were followed for 3–29 years with at least 2 visits separated by 3 or more years. The average rate of decline over the long-term can be described by an exponential function with a mean rate of loss of 8.1% of remaining cone function per year. The rate of decline can also be described by a spline model. Goodness of fit of the exponential model did not significantly differ from that of the spline model.

Figure 9

Data from Figure 8 replotted on a semi-log scale show that the rate of decline of cone ERG amplitudes from patients with retinitis pigmentosa can be described by a straight line even at low voltages, compatible with the idea that an exponential model can be used to describe the average long-term rate of loss of retinal function in patients with retinitis pigmentosa.

Figure 10

Full-field 30-Hz cone ERGs from a 27-year-old patient with retinitis pigmentosa without narrow bandpassed filtering and computer averaging (left) and with narrow bandpassed filtering and computer averaging (right). This patient had 20/20 vision and full visual fields with a V4e white test light.