This study answers a question posed by a previous study of cross-sectional data from patients with rhodopsin mutations in which we reported that disease severity at a given age varies with the location of a mutant amino acid residue within the opsin molecule.
4 In that study, for each measure of ocular function, the slope of the regression line representing ocular function versus age in patients with mutations affecting the intradiscal region was similar to that in patients with mutations affecting the cytoplasmic region, raising the possibility that the differences in severity may reflect differences in disease expression early in life and not in subsequent disease course. The present results show that differences in disease severity at a given age among patients with rhodopsin mutations reflect in large part differences in disease course. When we compared rates of change inferred from our cross-sectional data (based on first visits to our center of patients examined at different ages) to mean rates of change calculated from our longitudinal data, we found that the former significantly differed from the latter, pointing to the importance of longitudinal studies to measure disease course in individual patients.
We suspect that inferences about disease course based on cross-sectional data are flawed by ascertainment biases (e.g., on the part of patients seeking an evaluation when they first become symptomatic and on the part of investigators, as in this study, recruiting younger and older relatives, regardless of symptoms, because they were are found to carry a mutation). In the case of visual field progression, some patients who were ascertained for the first time late in life with small fields (see
Fig. 3 ) may never have had a normal field and were recruited regardless of symptoms because of our molecular genetics analyses. This ascertainment bias may have contributed to a steeper slope inferred from our cross-sectional data than what we measured based on our longitudinal follow-up. Because it is impossible to explain fully or understand these ascertainment biases that may have a profound effect on estimates of rates of change from cross-sectional data, the only way to estimate reliably a rate of decline over time in an individual patient is to observe the same patient over time, as we did in our longitudinal analyses.
Several previous studies have monitored the natural course of disease longitudinally in large cohorts of patients with retinitis pigmentosa. Unlike the present study, the sets of patients evaluated in those studies had unknown primary gene defects and probably were heterogeneous mixtures of patients with many of the dozens of genes that can cause retinitis pigmentosa. Based on those studies, the mean annual rate of decline of remaining visual acuity has been estimated at 1.0%,
18 3.7% (based on the individual data),
15 and 8.6%
19 in patients with retinitis pigmentosa in general and at 1.4% in patients with dominant retinitis pigmentosa.
19 Another study found that the median time to reach 20/200 was approximately 5 years in patients with a baseline visual acuity of 20/40.
20 Assuming that the mean would approximate the median, this corresponds to a 27% average annual rate of decline. Our rate of 1.8% in patients with rhodopsin mutations falls within the range of these values. The mean annual rate of decline of remaining visual field area has been estimated at 4.6%,
18 9.1%,
14 9.2% (based on the individual data),
15 and 13.5%.
16 Our rate of 2.6% is below the range of these values. The mean annual rate of decline of remaining cone ERG amplitude has been estimated at 13.3%
19 and 18.5%
18 in patients with retinitis pigmentosa in general and at 11.9% in patients with dominant disease.
19 Again, our rate of 8.7% is below these values. The results of the present study suggest that, on average, patients with rhodopsin mutations lose visual acuity at a rate similar to that observed in other patients with retinitis pigmentosa but lose peripheral retinal function, as monitored by visual field and full-field cone ERG testing, more slowly than other patients with retinitis pigmentosa.
Some of our patients ascertained in middle age had normal visual fields at baseline in response to the V4e test light (e.g., see
Fig. 5 ), consistent with the idea of a variable “critical age”
16 at which visual field loss begins with this test stimulus. However, we did not observe a ceiling effect in our cohort for change in visual field area and therefore did not exclude from longitudinal analysis patients with normal baseline visual fields or field data within the normal range. We similarly did not detect a ceiling effect for change in visual acuity or ERG amplitude, although we censored from longitudinal analysis visual acuities of 20/20 (except those that followed a lower value) because we did not code higher values. We excluded patients with rates of change that were identified as statistical outliers (who were as likely as not to show progression) and, as in prior studies,
11 14 15 19 patients with very low baseline values who, as a group, showed a floor effect. However, unlike some previous studies,
14 15 16 we included other patients who showed minimal or no progression, because we wanted our conclusions to be as representative as possible of patients with rhodopsin mutations and because our distributions for rate of change were Gaussian with these values included, thereby permitting us to use standard statistical methods to develop confidence limits for average rates of change and to compare rates of change between groups. It is likely that our observed mean rates of change for visual field and ERG progression were slower than those reported in some other studies, in part because of our inclusion of some patients who showed minimal or no progression.
In this study we grouped patients according to the regions of the rhodopsin molecule affected by sets of mutations. These regions are based on a three-dimensional structure of rhodopsin.
6 7 Our results showed that in groups of patients with rhodopsin mutations, disease course varies according to the region within the opsin molecule of an altered amino acid residue. The globular N terminus is thought to enable the proper folding of opsin and to confer stability to rhodopsin in the outer segment.
21 22 The plug appears to support the chromophore, 11-
cis retinal, in the pocket, while possessing some mobility to allow reconstitution of rhodopsin after bleaching.
7 The last five amino acids of the C terminus are necessary for proper sorting and vectorial transport of rhodopsin from the Golgi complex to the basal discs of the outer segment.
23 24 Transgenic mice, rats, and frogs with mutations affecting the C terminus have vesicles containing rhodopsin mislocalized to the inner segment cytoplasm and the plasma membrane or to the extracellular space adjacent to the border of the inner and outer segments.
25 26 27 The present study shows that patients with retinitis pigmentosa produced by rhodopsin mutations affecting the C terminus have more rapidly progressing disease with respect to visual field loss and ERG amplitude decline than patients with mutations affecting other regions. This suggests that impaired sorting and vectorial transport of rhodopsin are associated with a more rapid rate of photoreceptor degeneration than defects associated with impaired folding (i.e., the globule mutations) or the abnormal formation of a functional pocket for the binding of vitamin A (i.e., the plug mutations).
Although significant differences were observed in the average rate of change by the region of rhodopsin affected by a given mutation, we noted considerable variation in disease course among patients with mutations affecting the same region. Our regression model, which included age, gender, baseline function, and affected region of rhodopsin, accounted for only 20% to 34% of the variation in rates of change. This suggests that some factor(s) other than the gene defect itself—for example, modifier genes, diet, general health, or light exposure—may affect the clinical course of this condition. Knowledge of a patient’s rhodopsin mutation helps to predict the disease course, but most of the variation in rate of progression among these patients remains unexplained.