|Home | About | Journals | Submit | Contact Us | Français|
Provide an overview of some of our electroretinographic and psychophysical studies of normal development of rod function and their application to retinopathy of prematurity (ROP).
Electroretinographic (ERG) responses to full-field stimuli were recorded from dark adapted subjects. Rod photoreceptor sensitivity, SROD, was calculated by fit of a biochemical model of the activation of phototransduction to the ERG a-wave. Dark adapted psychophysical thresholds for detecting 2° spots in parafoveal (10° eccentric) and peripheral (30° eccentric) retina were measured and the difference between the thresholds, Δ10-30, was examined as a function of age. SROD and Δ10-30 in term born and former preterm subjects were compared.
In term born infants, (1) the normal developmental increase in SROD changes proportionately with the amount of rod visual pigment, rhodopsin, and (2) rod mediated function in central retina is immature compared to that in peripheral retina. In subjects born prematurely, deficits in rod photoreceptor sensitivity persist long after active ROP has resolved. Maturation of rod mediated thresholds in the central retina is prolonged by mild ROP.
Characterization of the development of normal rod and rod mediated function provides a foundation for understanding ROP.
The retina, an accessible part of the nervous system in which sensory and neural processes can be studied using non-invasive procedures, is immature at birth and continues to develop postnatally. We have tested hypotheses about the normal immature neurosensory retina in infants and children.1-23 In this paper, we present two main concepts about the development of normal scotopic function that have emerged from these studies and their application to one pediatric retinal disorder, retinopathy of prematurity (ROP).
One concept concerns immaturity of the rod photoreceptors. At term birth, all retinal layers are present and the number of retinal cells remains about constant through ensuing years.24 Thus, our early studies that documented differences between infant and adult electroretinographic (ERG) responses to full-field stimuli suggested immaturities of the retinal cells.1, 25, 26 The ERG data, as well as psychophysical data,7, 18, 21 identified the rod photoreceptor as a site of critical immaturity in young infants. This led us to determine the growth curve for the rod visual pigment, rhodopsin,27-29 and to study phototransduction processes in infants.6, 20 We found that the amount of rhodopsin increases during development 28, 29 and that rhodopsin content scales the kinetics of activation and deactivation of phototransduction in dark adapted infants.6, 19 We have backed up these conclusions about development of human rods with studies of rat rods.30-35
The second concept is that development of central retinal function is delayed relative to that in more peripheral retina. Delayed development of the central retina pertains to both rod 12, 17, 36 and cone 22 mediated function; this paper considers rod function only. In keeping with the delayed developmental elongation of the parafoveal rod outer segments,36-38 psychophysically determined parafoveal thresholds of young infants are elevated relative to peripheral thresholds.12, 17, 36
Herein we provide an overview of some electroretinographic and psychophysical studies of the development of normal rod and rod-driven function. We have applied our measurement and analytical procedures to investigation of pediatric retinal disorders 39-43 and illustrate this by presenting some data obtained in ROP.44-49 ROP is the most common pediatric retinal disease to cause life-long, bilateral visual impairment.50
The coincidence of age at onset of ROP 51 and acceleration of the rhodopsin growth curve (Figure 1) alerted us to the pivotal role of the immature rods in the ROP disease process. Specifically, it is not until the age at which rapid increase in rhodopsin content and developmental elongation of the rod photoreceptor outer segments is in progress that ROP makes its clinical appearance, which is marked by the formation of abnormal retinal blood vessels.51 We theorized that rods have a pivotal role in ROP because, on one hand, the developing rods' escalating demand for aerobic energy contributes to the relative hypoxia of the neural retina that, in turn, drives development of the abnormal retinal vasculature.52 On the other hand, incompletely met energy demands can injure the developing rods.53-55 We hypothesized that (1) such injury would cause rod dysfunction in ROP subjects, and (2) the late maturing rods in the central retina would be particularly vulnerable to the effects of ROP, even if mild. We have tested these hypotheses using electroretinographic and psychophysical procedures.
We recorded ERG responses to full-field stimuli in dark adapted, awake infants using a corneal contact electrode (Figure 2A). The full-field ERG evaluates the function of the retina as a whole. ERG responses to a several log unit range of stimulus intensities were recorded and a-waves, which represent rod photoreceptor activity, were analyzed using mathematical models.56-59 In these models, SROD is a sensitivity parameter related to the gain of the activation of phototransduction. It is based on the time constants of the molecular processes that are initiated with photon absorption by rhodopsin.58, 59
In normal term born infants, we found that the growth curve for SROD 6 is similar to that for rhodopsin.28 In infants and children born prematurely and tested at post term ages, we found ERG evidence of persistent rod dysfunction long after active ROP disease had resolved (Figure 3).46, 48 In the majority of subjects with mild, untreated ROP (30 of 40), SROD was below average; by clinical criteria, their ROP had resolved without residua. Of the 11 subjects with more severe ROP treated by laser ablation of the peripheral avascular retina at preterm ages, seven had SROD values below the 95% prediction limit for normal; the area of ablated retina did not account for the deficits in SROD. None of the ROP subjects had retinal detachments. In the subjects who never had ROP (N=17), SROD values were distributed around the average normal value.
In contrast to ERG responses to full-field stimuli in which spatial information is lacking, behavioral responses to stimulation of small patches of retina can be studied using psychophysical procedures (Figure 2B). The area of central retina with delayed rod outer segment development is too small (<5% of the total retinal area) to have a detectable impact on ERG responses. We devised a psychophysical procedure in which the dark adapted threshold in parafoveal retina was compared to the threshold in peripheral retina.36 Stimulus conditions were selected such that in adults, thresholds are equal at the two sites; that is, delta 10-30 (Δ10-30) in adults is zero.
We used a two-alternative, spatial forced-choice preferential looking procedure to estimate dark adapted thresholds in 2° diameter patches of parafoveal (10° eccentric) and peripheral (30° eccentric) retina (Figure 4). We presented test stimuli on the right or left of a rear projection screen. A fix and flash paradigm was used to control the retinal site stimulated.61 An adult observer, unaware of the position of the test spot, reported the direction of the infant's looking behavior, right or left. For speed, and thus feasibility for evaluating infants, we used a rapid, up-down staircase.62 In basic work, we have shown that thresholds obtained using the staircase are in good agreement with those obtained using the method of constant stimuli.7
In healthy term born 10 week olds, we found that the parafoveal threshold was more elevated than the peripheral threshold. The course of maturation was such that threshold at both sites reached adult values by age 6 months.12, 17
We used the Δ10-30 procedure to test the hypothesis that ROP alters the course of development of the rods in the late maturing central retina. We found that in infants with a history of mild, untreated ROP which had resolved completely and spontaneously, the course of maturation of the thresholds was prolonged (Figure 5A) compared to that in normal term born infants.44 In former preterms who never had ROP (Figure 5B), the course of maturation of the thresholds was indistinguishable from that in term born infants.
The results shown in Figures 3 and and55 illustrate application of some of the methods that we have used in studies of normal retinal development to ROP. Long after active ROP has resolved, deficits in rod photoreceptor sensitivity persist (Figure 3). In the central retina, maturation of rod mediated thresholds is prolonged by mild ROP (Figure 5A). In rat models of ROP, the fine structure of the rods is altered 54 and SROD is attenuated53, 55, 63 in a fashion similar to that in human ROP subjects (Figure 3). Furthermore, in rat models, rod dysfunction antedates 55 and predicts 53 subsequent ROP vascular status, and early pharmaceutical intervention that targets fundamental rod function improves vascular outcome.64 These preclinical data offer a new and promising perspective on treatment of human ROP.
The dark adapted rod photoreceptors do not tell the whole story about retinal function in normal development or in ROP. Scotopic ERG and psychophysical data provide evidence that the postreceptor retina re-organizes both in normal development and during the evolution and resolution of ROP.14-16, 18, 21, 23, 47, 48, 65 Cone photoreceptor and cone-driven postreceptor ERG responses to full-field stimuli are also immature, but in keeping with their earlier anatomic development,24 cones are relatively more mature than rods 20 and are less affected by ROP;45 possibly earlier maturation is protective. Cone-driven postreceptor function in the central retina is remarkably immature in young term born infants 22 and is altered by mild ROP in a fashion that forecasts 66 structural changes in the neurovascular retina, which we have demonstrated using high resolution imaging techniques.67
More remains to be learned about both scotopic and photopic processes in the normal, immature mammalian retina. For instance, retinal adaptation in steady state conditions 18, 21, 47 and during recovery from light exposure 8, 26 warrant further study. ERG and psychophysical investigations in the living eye in animal models and in patients will surely continue to identify disease processes and guide the design of novel treatments.
Supported by National Institutes of Health grant R01-EY10597.