Baseline and Constriction Rate Associations
After these aforementioned evaluations, data from right and left eyes were combined for further analyses. The association between constriction rates with baseline measurements and age is shown in Table 4. Ring size at baseline as quantified with all three outer border metrics correlated negatively and strongly with age (r ≥ −0.7674, P < 0.0001 for all 3). Correlation between constriction rates and baseline size for each of the three metrics was found to be strongest for the area-derived (r = 0.8502, P < 0.0001) and vertical diameter–derived metrics (r = 0.5969, P < 0.0001). Correlation for the horizontal diameter–derived metric was also significant, albeit weaker (r = 0.3554, P = 0.0012). In addition, area-derived constriction rates showed the strongest negative correlation with age (r = −0.6436, P < 0.0001), whereas that of vertical diameter rates were moderate in strength (r = −0.4145, P = 0.0001).
A two-way ANOVA was performed to investigate the effects of age, genotype, and interaction between age and genotype on constriction rates. Analysis of variance results together with mean ± SD values for constriction rates in each category are given in Table 5. The effect of age was significant for constriction rates derived from ring area (P < 0.0001) and vertical diameter (P = 0.0002). Post hoc pairwise comparisons between age categories for area-derived constriction rates disclosed significant differences between younger and older subjects, that is, age Category 1 (youngest) versus all other age categories (P < 0.0001); Category 2 versus Category 5 (P = 0.0072); and Category 3 versus Category 5 (P = 0.0030). Similar findings were made with post hoc pairwise comparisons for vertical diameter–derived constriction rates, in that differences between Category 1 versus Category 5 (P = 0.0002) and Category 1 versus Category 4 (P = 0.0064) reached statistical significance. The effect of genotype on constriction rates, by contrast, was insignificant.
In this article, we present the first comprehensive FAF study to describe bilateral progression of disease in a cohort comprising entirely of molecularly proven subjects with RPGR-associated RP. Forty-eight percent of subjects in our study had rings, in comparison with 55% to 69% described in other RP cohorts.15,19,23,24,29 Our subjects with rings are younger in age (16.3 ± 7.9 years of age at time of baseline imaging) in comparison with mean ages of 35 to 49 years described in previous studies.12,18,22,23,29 However, these studies were composed of subjects with both ADRP and ARRP and thereby are genetically diverse, with the possible exception of a study investigating Usher syndrome Type I and II—although these are also genetically heterogeneous.19 Only one study included subjects with molecularly confirmed XLRP—even then comprising only 6% of the entire cohort.24 It is therefore unsurprising that ring findings (including lack/loss of an FAF ring in older subjects) and other FAF observations that signify more advanced disease at a younger age have been identified in our cohort of molecularly proven RPGR RP subjects, as it is relatively more severe with faster progression compared with other genotypes.
Correlation Between Age and Baseline Measurements
We found strong and significant negative correlations between age and baseline measurements for all three outer border ring metrics (r ≥ −0.7674, P < 0.0001 for all 3 metrics). Our findings substantiate weaker associations previously reported between age and ring size in cohorts of mixed inheritance.12,17,30 In addition, numerical age can arguably serve as a good approximation for disease duration in RPGR RP given the early onset of disease in childhood typical for the condition.9
The progression rates described in our study reflect our cohort of molecularly proven RPGR subjects, which overall renders a greater disease severity. This is in contrast with previous studies where cohorts compose of subjects with mixed inheritance and RP of lesser severities. Thus, our mean annual area of autofluorescence constriction rate of 1.42 mm2 (10.7%), mean annual horizontal and vertical diameter constriction rates of 214.1 μm (5.5%) and 207.9 μm (6.3%), respectively, are greater than previously reported.17–19,24 One recent study of 71 subjects where two-thirds had ARRP, 27% with ADRP, and only 6% with XLRP found mean annual constriction rates of 147 μm (4.1%) and 121 μm (4.0%) for horizontal and vertical diameter outer border ring metrics, respectively.24 A study comprising 8 eyes with ADRP and 6 with ARRP (with no XLRP subjects) reported lower annual outer ring rates of constriction of 2.5% and 2.1% for horizontal and vertical diameters, respectively.18 A mean annual ring constriction rate of 60 μm (3.0%) calculated from outer border ring radius measurements (4.0% with inner border ring radius) was found in 13 patients with Usher syndrome.19 Robson et al described 30 RP subjects with a mixed mode of inheritance who underwent serial FAF imaging with a mean follow-up of 4 years from baseline. Progression was seen in 17 subjects, 16 of these had ARRP including Usher syndrome, with a mean annual inner border ring radius constriction rate of 5.6%.17
Overall Exponential Rate of Decline
To the best of our knowledge, this is the first study to demonstrate an exponential rate of decline in RP on the basis of serial FAF imaging, with the implication that the rate of progression decreases with time (Figure 4). Half-lives of outer border–derived ring area calculated separately for right and left eyes are similar at 6 years. Phenotypic heterogeneity in XLRP-RPGR, however, necessitates individual observations.9 An exponential decline has previously been demonstrated in functional studies,31–33 which were genetically heterogenous with the exception of Iannaccone et al32 who studied subjects with Usher syndrome Type II (which nevertheless was also genetically heterogenous, as subjects were not genetically confirmed). Massof et al observed that the visual field loss occurred exponentially with a similar level of decline across all forms of RP, and this was believed to occur secondary to a common and final process of retinal degeneration. They extrapolated data to estimate the age at which visual field loss began and hypothesized that any difference was due to the earlier onset in XLRP.31
We did not set out to directly compare progression between XLRP-RPGR and other forms of RP in our study; however, we note that others have found a greater rate of progression in XLRP. For instance, Birch et al34 found that the inheritance pattern in RP had a significant effect on progression rates, with annual rod electroretinogram threshold elevation being highest in XLRP and lowest in ADRP. Sandberg et al35 observed a greater annual decline in visual field area in subjects with RPGR-associated RP, compared with subjects with RHO-associated autosomal dominant RP. A 4.7% mean annual exponential rate of decline was observed for RPGR-associated RP, compared with 2.9% for RHO-associated RP. Using structural optical coherence tomography measurements of ellipzoid zone width, Cai et al36 found a significantly greater rate of progression in subjects with XLRP (9.6%/year) compared with ADRP (3.4%/year).
Correlation Between Age and Progression
We found a strong and significant negative correlation between age and rate of area constriction (r = −0.6436, P < 0.0001). Correlation between age and rate of vertical diameter constriction was moderate (r = −0.4145, P = 0.0001), whereas correlation between age and horizontal diameter constriction was weak and insignificant after Bonferroni correction (r = −0.2366, P = 0.0346). By contrast, Robson et al17 did not find correlation between age and constriction rates, nor did they find correlation between constriction rates and baseline ring size. Sujirakul et al24 also did not find differences in constriction rates with age. Both studies comprised a heterogenous mix of RP subjects with various forms of inheritance patterns. This was alluded to as a reason for noncorrelation by Robson et al.17 It is thus possible that potential correlation with age may have been affected by the averaging of results secondary to the inclusion of subjects with different forms of RP, given the wide age range of disease onset and progression rates between different forms.2
Correlation Between Baseline Ring Metrics and Progression
Correlation between baseline size with respective constriction rates was all significant and positive. Area constriction rate correlated very strongly with baseline area (r = 0.8502, P < 0.0001), whereas vertical diameter rate constriction demonstrated a strong correlation with vertical diameter at baseline (r = 0.5969, P < 0.0001). Correlation was weakest for horizontal diameter (r = 0.3554, P = 0.0012). This finding of a significant association between larger baseline measurements and faster progression rates was also reported by Sujirakul et al24 when the cohort was divided into subjects with greater or lesser than a baseline vertical diameter of 3,000 μm. Interestingly, the association between horizontal diameter constriction and baseline metrics in their study was not significant, in keeping with our finding of ring horizontal diameter having the weakest correlation of the three metrics.
Effects of Age and Genotype on Progression
The effects of age were significant on both ring area and vertical diameter constriction rates. Post hoc comparisons showed significant differences between younger subjects, in particular those in age Category 1 compared with older age categories, indicating that progression is maximal in the youngest subjects. There was no effect of genotype on constriction rates.
Although speculative in nature, the effects of age could perhaps be explained by a biochemical model hypothesized by Clarke et al37 whereby each photoreceptor has a risk of death that is constant over time and occurring at random. In the case of younger subjects with larger rings, these large rings represent greater numbers of surviving photoreceptors. It is conceivable that more photoreceptors will die off at the early stages because of the greater numbers of photoreceptors present. This would give rise to a greater initial rate of decline. With age, as the numbers of surviving photoreceptors decrease, so will there be a concomitant decrease in the numbers of photoreceptors dying. Thus, the rate of decline slows in later age. This model can account for the exponential graph shown in Figure 4.
A high level of interocular symmetry at baseline and for overall rates of progression has been demonstrated in our study. Interocular symmetry at baseline has also been reported by others, for example by Robson et al11 who described a strong interocular correlation (r = 0.94) between internal radii of the FAF rings, as well as symmetry in electrophysiological testing (r = 0.94). More recently, Sujirakul et al30 also reported good overall interocular symmetry of vertical and horizontal diameter ring measures (r = 0.99 and 0.98, respectively). In addition, they did not find a difference in progression rates between eyes.24
Wakabayashi et al assessed interobserver variability for horizontal and vertical ring diameter metrics (not stated whether these were inner or outer ring measurements) and their coefficient of repeatability was 240 μm for horizontal and 250 μm for vertical diameter. The 95% LOA was −230 μm to 270 μm and −230 μm to 280 μm. Test–retest variability was 14.0% for horizontal diameter and 20.3% for vertical diameter.16 Three observers in Sujirakul et al measured both horizontal and vertical outer ring diameters. Limits of agreement pairs were calculated for each metric by comparing two observers at a time. The greatest value from the LOA pairs was used as a cutoff for measurement error, that is, measurements exceeding this cutoff were deemed clinically significant. Here, 421 μm and 412 μm were the cutoffs for vertical and horizontal diameter measurements, respectively.30
Sujirakul et al24 also assessed intraobserver variability—arguably more important for planned Phase I/II RP clinical trials. Test–retest variability was 9.5% for horizontal and 9.6% for vertical outer ring diameter metric. Annual rate of constriction in their study was 4.1% (147 μm/year) and 4.0% (121 μm/year) for horizontal and vertical diameter ring metrics, respectively. The 95% LOA calculated from published data was −137 μm to 336 μm for horizontal diameter and −142 μm to 316 μm for vertical diameter metrics. Our intraobserver variability is comparable, with our test–retest variability of 9.2% for horizontal diameter and 11.0% for vertical diameter metrics, and 95% LOA of −249 μm to 347 μm and −230 μm to 326 μm, respectively.
These studies solely used ring diameter metrics for which test–retest variabilities were twice the constriction rate.16,24,30 We have identified ring area (derived from outer border) to be a robust metric—as despite a marginally higher test–retest variability when compared with the diameter metrics, annual rate of constriction is greatest when quantified with the area metric, and approximately of the same magnitude as its corresponding test–retest variability (Table 1). We are confident that constriction rates obtained in our study are robust and real, as multiple observations with relatively long follow-up were obtained on our subjects, in addition to the robust method used to derive constriction rates.
Assessing Suitability by Comparing Annual Mean Constriction Rates With Largest 95% Limits of Agreement Value
Mean rate of area constriction for our entire cohort was 1.42 mm2/year, which is greater than the larger 95% LOA value of 1.11 mm2. In direct contrast, the mean rate of outer border horizontal diameter constriction was 214.1 μm/year that is smaller than the larger LOA value of 347 μm. Likewise, vertical diameter constriction was 207.9 μm/year, which again is smaller than the larger LOA value of 326 μm. As shown in Table 5 and Figure 5, constriction rates are greatest in younger subjects with those in age Category 1 having a mean area constriction rate of 4.33 mm2/year. Thus, our findings of ring area as being the metric of choice to quantify progression may arguably be even more sensitive and robust in younger patients with inherently faster rates of progression, who moreover will likely constitute the target group for intervention.
Choice of Ring Metrics
Previous groups have used various FAF metrics for analysis with no apparent consensus. These include measurements of ring area,15,23 ring horizontal diameter,15,16,18,24 ring vertical diameter,16,18,24,29 and ring radius.11,12,17,19,22 Others have preferred either measurements taken from the inner borders of the rings15 or from the outer borders,18,24,30 with the rationale that one border can be more accurately demarcated over the other. It is of note that none of these studies have provided results from objective investigations of measurement repeatability to substantiate their preferences. We have undertaken such assessments for ring and diameter metrics derived from both borders in our study. Furthermore, none of the studies that measure diameters on sequential images have corrected for image rotation, which represents a significant limitation.
Our study with a relatively long period of follow-up has allowed us to obtain sufficient data points to be the first FAF study to demonstrate an overall exponential decline in progression rate with age. Future studies will likely require longer periods of follow-up that is, 20 years or more to plot individual exponential decline.
Fundus autofluorescence imaging is reproducible and widely available in retinal clinics worldwide; hence, we advocate its regular use in monitoring progression in RP. Our finding of outer border–derived ring area as the most sensitive and valid metric to detect change will guide metric choice for interventional trials when assessing disease progression.
In addition, we have presented longitudinal data using FAF ring metrics to characterize baseline values and progression rates in RPGR-associated RP. The rate of progression is dependent on age and baseline ring size. In general, there is good overall interocular symmetry. These findings will be useful in informing patient selection and outcome measures in future treatment trials, as well as clinicians providing prognostic information to patients with RPGR-associated RP.
1. Hartong DT, Berson EL, Dryja TP. Retinitis pigmentosa
. Lancet 2006;368:1795–1809.
2. Hamel C. Retinitis pigmentosa
. Orphanet J Rare Dis 2006;1:40.
3. Xu Y, Guan L, Shen T, et al. Mutations of 60 known causative genes in 157 families with retinitis pigmentosa
based on exome sequencing. Hum Genet 2014;133:1255–1271.
4. You QS, Xu L, Wang YX, et al. Prevalence of retinitis pigmentosa
in North China: the Beijing eye public health care project. Acta Ophthalmol 2013;91:e499–e500.
5. Haim M. Epidemiology of retinitis pigmentosa
in Denmark. Acta Ophthalmol Scand Suppl 2002;233:1–34.
6. Pelletier V, Jambou M, Delphin N, et al. Comprehensive survey of mutations in RP2 and RPGR
in patients affected with distinct retinal dystrophies: genotype-phenotype correlations and impact on genetic counseling. Hum Mutat 2007;28:81–91.
7. Shu X, Black GC, Rice JM, et al. RPGR
mutation analysis and disease: an update. Hum Mutat 2007;28:322–328.
8. Sharon D, Sandberg MA, Rabe VW, et al. RP2 and RPGR
mutations and clinical correlations in patients with X-linked retinitis pigmentosa
. Am J Hum Genet 2003;73:1131–1146.
9. Tee JJ, Smith AJ, Hardcastle AJ, Michaelides M. RPGR
-associated retinopathy: clinical features, molecular genetics, animal models and therapeutic options. Br J Ophthalmol 2016;100:1022–1027.
10. von Ruckmann A, Fitzke FW, Bird AC. Distribution of pigment epithelium autofluorescence in retinal disease state recorded in vivo and its change over time. Graefes Arch Clin Exp Ophthalmol 1999;237:1–9.
11. Robson AG, El-Amir A, Bailey C, et al. Pattern ERG correlates of abnormal fundus autofluorescence
in patients with retinitis pigmentosa
and normal visual acuity. Invest Ophthalmol Vis Sci 2003;44:3544–3550.
12. Popovic P, Jarc-Vidmar M, Hawlina M. Abnormal fundus autofluorescence
in relation to retinal function in patients with retinitis pigmentosa
. Graefes Arch Clin Exp Ophthalmol 2005;243:1018–1027.
13. Robson AG, Saihan Z, Jenkins SA, et al. Functional characterisation and serial imaging of abnormal fundus autofluorescence
in patients with retinitis pigmentosa
and normal visual acuity. Br J Ophthalmol 2006;90:472–479.
14. Robson AG, Michaelides M, Saihan Z, et al. Functional characteristics of patients with retinal dystrophy that manifest abnormal parafoveal annuli of high density fundus autofluorescence
; a review and update. Doc Ophthalmol 2008;116:79–89.
15. Aizawa S, Mitamura Y, Hagiwara A, et al. Changes of fundus autofluorescence
, photoreceptor inner and outer segment junction line, and visual function in patients with retinitis pigmentosa
. Clin Exp Ophthalmol 2010;38:597–604.
16. Wakabayashi T, Sawa M, Gomi F, Tsujikawa M. Correlation of fundus autofluorescence
with photoreceptor morphology and functional changes in eyes with retinitis pigmentosa
. Acta Ophthalmol 2010;88:e177–e183.
17. Robson AG, Tufail A, Fitzke F, et al. Serial imaging and structure-function correlates of high-density rings of fundus autofluorescence
in retinitis pigmentosa
. Retina 2011;31:1670–1679.
18. Lima LH, Burke T, Greenstein VC, et al. Progressive constriction of the hyperautofluorescent ring in retinitis pigmentosa
. Am J Ophthalmol 2012;153:718–727. 727.e711–e712.
19. Fakin A, Jarc-Vidmar M, Glavac D, et al. Fundus autofluorescence
and optical coherence tomography in relation to visual function in Usher syndrome type 1 and 2. Vis Res 2012;75:60–70.
20. Robson AG, Lenassi E, Saihan Z, et al. Comparison of fundus autofluorescence
with photopic and scotopic fine matrix mapping in patients with retinitis pigmentosa
: 4- to 8-year follow-up. Invest Ophthalmol Vis Sci 2012;53:6187–6195.
21. Lenassi E, Troeger E, Wilke R, Hawlina M. Correlation between macular morphology and sensitivity in patients with retinitis pigmentosa
and hyperautofluorescent ring. Invest Ophthalmol Vis Sci 2012;53:47–52.
22. Greenstein VC, Duncker T, Holopigian K, et al. Structural and functional changes associated with normal and abnormal fundus autofluorescence
in patients with retinitis pigmentosa
. Retina 2012;32:349–357.
23. Iriyama A, Yanagi Y. Fundus autofluorescence
and retinal structure as determined by spectral domain optical coherence tomography, and retinal function in retinitis pigmentosa
. Graefes Arch Clin Exp Ophthalmol 2012;250:333–339.
24. Sujirakul T, Lin MK, Duong J, et al. Multimodal imaging of central retinal disease progression in a 2-year mean follow-up of retinitis pigmentosa
. Am J Ophthalmol 2015;160:786–798 e784.
25. Delori FC, Dorey CK, Staurenghi G, et al. In vivo fluorescence of the ocular fundus exhibits retinal pigment epithelium lipofuscin characteristics. Invest Ophthalmol Vis Sci 1995;36:718–729.
26. Feeney-Burns L, Berman ER, Rothman H. Lipofuscin of human retinal pigment epithelium. Am J Ophthalmol 1980;90:783–791.
27. Dorey CK, Wu G, Ebenstein D, et al. Cell loss in the ageing retina: relationship to lipofuscin accumulation and macular degeneration. Invest Ophthalmol Vis Sci 1989;30:1691.
28. von Ruckmann A, Fitzke FW, Bird AC. Distribution of fundus autofluorescence
with a scanning laser ophthalmoscope. Br J Ophthalmol 1995;79:407–412.
29. Murakami T, Akimoto M, Ooto S, et al. Association between abnormal autofluorescence and photoreceptor disorganization in retinitis pigmentosa
. Am J Ophthalmol 2008;145:687–694.
30. Sujirakul T, Davis R, Erol D, et al. Bilateral concordance of the fundus hyperautofluorescent ring in typical retinitis pigmentosa
patients. Ophthalmic Genet 2015;36:113–122.
31. Massof RW, Dagnelie G, Benzschawel T. First order dynamics of visual field loss in retinitis pigmentosa
. Clin Vis Sci 1990;5:1–26.
32. Iannaccone A, Kritchevsky SB, Ciccarelli ML, et al. Kinetics of visual field loss in Usher syndrome type II. Invest Ophthalmol Vis Sci 2004;45:784–792.
33. Berson EL. Long-term visual prognoses in patients with retinitis pigmentosa
: the Ludwig von Sallmann lecture. Exp Eye Res 2007;85:7–14.
34. Birch DG, Anderson JL, Fish GE. Yearly rates of rod and cone functional loss in retinitis pigmentosa
and cone-rod dystrophy. Ophthalmology 1999;106:258–268.
35. Sandberg MA, Rosner B, Weigel-DiFranco C, et al. Disease course of patients with X-linked retinitis pigmentosa
due to RPGR
gene mutations. Invest Ophthalmol Vis Sci 2007;48:1298–1304.
36. Cai CX, Locke KG, Ramachandran R, et al. A comparison of progressive loss of the ellipsoid zone (EZ) band in autosomal dominant and x-linked retinitis pigmentosa
. Invest Ophthalmol Vis Sci 2014;55:7417–7422.
37. Clarke G, Collins RA, Leavitt BR, et al. A one-hit model of cell death in inherited neuronal degenerations. Nature 2000;406:195–199.
Keywords:© 2018 by Ophthalmic Communications Society, Inc.
fundus autofluorescence; retinal diseases; retinitis pigmentosa; RPGR