First order dynamics of visual field loss in retinitis pigmentosa

1. Typical retinitis pigmentosa (RP) can be subdivided according to mode of inheritance and according to the relation of rod sensitivity loss to cone sensitivity loss (type 1 and type 2). Although clinical impression suggests that there are differences among genetic types in the RP progression, systematic natural history data have been sparse and not definitive. 2. The present study reports an analysis of the dynamics of prospectively-measured visual field loss in 172 typical RP patients tested repeateedly over periods ranging from 2.5 to 10 yr. Visual fields were measured using a kinetic method on the Goldmann perimeter with the Goldmann V/4e and II/4e test stimuli. 3. Visual fields were transformed to retinal coordinates using the method of Drasdo and Fowler (1974). Visual field area in mm² of retina was computed for each measured field. The average normal visual field areas, based on measurements from 155 healthy volunteers, are 2.87 log mm² for the V/4e stimulus (SD = 0.034 log unit) and 2.82 log mm² for the II/4e stimulus (SD = 0.056 log unit). 4. For RP patients, the rate of change of visual field area was found to be proportional to visual field area, whereas the normalized rate of change was independent of visual field area. Thus, to the first order, visual field area is lost in RP according to an exponential decy function. 5. The average time constants (years to 1/e loss in visual field area) are 8.4 yr (SD = 4.9 yr) for the V/4e stimulus and 7.4 yr (SD = 4.7 yr) for the II/4e stimulus. Although relatively high confidence can be placed in this difference (p = 0.055), we found that the difference between time constants for the two stimuli does not reflect different dynamics, rather it could be attributed to departures from the first order exponential function for the largest visual field areas. When the V/4e distribution was limited to visual field areas of 500 mm² or less (the same range as the II/4e data), the time constants for the two simuli were indistinguishable (7.8 ± 4.6 yr for the truncated V/4e distribution). 6. There were no significant differences among time constants across genetic subtypes. The time constants for type 1 RP were significantly longer than the time constants for type 2 RP (2.2 yr difference for the II/4e distributions and 3.4 yr difference for the truncated V/4e distributions). 7. Critical ages, i.e. extrapolated age of visual field loss onset, were computed for each RP patient by assuming the average time constant and the average normal visual field area. The average critical ages for all RP were 28 yr (SD = 13.9 yr) for the V/4e stimulus and 22 yr (SD = 13.8 yr) for the II/4e stimulus. The V/4e and II/4e critical ages were proportionally related (r = 0.96; m = 1.03, b = 6 yr). Critical age was proportional to the age of first visual field loss symptoms reported by the patient (r = 0.68; m = 0.99, b = -3.27 yr for the V/4e distribution and r = 0.67; m = 1.01, b = 2.48 yr for the II/4e distribution). 8. X-linked RP patients exhibited the earliest average critical age (P = 0.088 for the V/4e distribution and P = 0.057 for the II/4e distribution). The average critical age for type 1 RP was less than the average critical age for type 2 RP (P = 0.07 for both the V/4e and II/4e distributions). 9. Further analysis of the RP natural history data of Berson et al. (1985) and the RP natural history data of Sunga and Sloan (1967) demonstrated that their results, albeit not their conclusions, are in point-for-point agreement with the results and conclusions of the present study. 10. By modeling the progressive visual field loss of RP as the product of a sensitivity loss gradient, moving radially according to the exponential decay function, and an initial retinal sensitivity profile, the difference between time constant distributions for type 1 and type 2 RP can be attributed to the previously reported difference in initial retinal sensitivity profiles between the two RP subtypes. The conclusion of this analysis is that the first order dynamics of the progressive retinal degeneration is the same in all types of RP. 11. We propose that the retinal degeneration of RP is a secondary phenomenon, a final common pathway for a variety of retinal diseases. It appears that the RP diagnosis and defining clinical characteristics can be attributed to this secondary phenomenon with its own internal driving forces and kinetics. We suggest that a likely candidate for the progressive retinal degeneration of RP is pigment epithelial cell migration and proliferation.