Objective To determine the refractive error in patients with autosomal recessive retinitis pigmentosa (arRP) caused by RP1 mutations and to compare it with that of other genetic subtypes of RP.
Methods Twenty-six individuals had arRP with RP1 mutations, 25 had autosomal dominant RP (adRP) with RP1 mutation, 8 and 33 had X-linked RP (xlRP) with RP2 and RPGR mutations, respectively, 198 and 93 had Usher syndrome and arRP without RP1 mutations, respectively. The median of the spherical equivalent (SE) and the IQR (Q25–Q75) was determined and multiple comparisons were performed.
Results arRP patients with RP1 mutations had SE median at −4.0 dioptres (D) OD (Ocula Dextra); −3.88 D OS (Ocula Sinistra), whereas arRP patients without RP1 mutations (−0.50 D OD; −0.75 D OS) and Usher syndrome patients (−0.50 D OD; −0.38 D OS) were significantly less myopic (p<0.0001). Conversely, myopia of xlRP patients with either an RPGR mutation (−4.50 D OD; −5.25 D OS) or an RP2 mutation (−6.25 D OD; −6.88 D OS) was not significantly different from the arRP group with RP1 mutations. arRP without RP1 mutations, Usher syndrome and adRP with RP1 mutation had a narrow IQR (−9.06 to −1.13 D), whereas arRP with RP1 mutations and xlRP with RP2 or RPGR mutations had a larger range (−9.06; −1.13 D).
Conclusions arRP patients with RP1 mutations have myopia not different from patients with xlRP with RP2 or RPGR mutations, while RP patients from other genetic subgroups were emmetropic or mildly myopic. We suggest that arRP patients with high myopic refractive error should be preferentially analysed for RP1 mutations.
Statistics from Altmetric.com
Retinitis pigmentosa (RP) or rod-cone dystrophy is the most frequent form of inherited retinal dystrophies, featuring night blindness, progressive loss in the peripheral visual field and decrease in visual acuity, leading to total blindness after several decades of evolution.1 Fundus examination shows characteristic bone spicule-shaped pigment deposits in the peripheral retina with attenuation of retinal vessels and pallor of the optic disc. The prevalence of RP is approximately 1/3500 to 1/4000 in developed countries and the mode of inheritance can be autosomal dominant (30%–40%), autosomal recessive (50%–60%) or X-linked (5%–15%). Non-syndromic RP is genetically highly heterogeneous with 69 disease-causing genes currently known (http://www.sph.uth.tmc.edu/retnet), including 24 genes causing autosomal dominant forms, 42 genes involved in autosomal recessive inheritance and three genes responsible for X-linked forms. Besides non-syndromic forms, there are also syndromic RPs, the most frequent being Usher syndrome, which associates RP with congenital deafness and accounts for about 15% of all RP cases.2
Refractive errors are frequently reported in inherited retinal dystrophies. Remarkably, in the congenital forms such as Leber congenital amaurosis, there is frequent high hyperopia, from +6 to +12, particularly in those very early forms due to mutations in GUCY2D, RPGRIP1, CRX or CEP290.3 In contrast, in retinal dystrophies with apparent onset several years after birth or later such as in RP, there is a significant skew of the mean refractive error towards moderate myopia and astigmatism,4 with a mean spherical equivalent (SE) of −1.86 in RP in comparison to +1.00 as found in a large healthy population.5 Strikingly, it was shown that patients with xlRP had a significantly higher myopic refractive error than patients with other forms of RP, reaching a mean SE of −5.51 vs −1.20 for non-xlRP.5 Another study also observed that xlRP patients had −2.00 or greater myopia,6 and myopia was otherwise consistently associated with xlRP in many studies.7 ,8
RP1 is one of the most frequently mutated genes in autosomal dominant forms of RP, accounting for 3%–6% of all cases.9 In the last decade, RP1 mutations were also reported in autosomal recessive forms of RP,10–21 suggesting that it might also be one of the most prevalent genes in this type of inheritance. While examining clinical features of recessive RP patients with RP1 mutations, we observed that they had high myopic refractive error.15 In this study, we clinically characterised 26 RP patients with recessive RP1 mutations and found that most of them indeed had myopia, being not different from xlRP due to RP2 or RPGR mutations in terms of refractive errors.
Materials and methods
Patients, refraction and visual acuity
Informed consent was obtained from each individual for blood withdrawal and clinical study. The study was done under the authorisation of the Department of Ophthalmology of the Hospital of Montpellier # 11018S from the French Ministry of Health for biomedical research in the field of physiology, pathophysiology, epidemiology and genetics in ophthalmology. A total of 26 RP patients who were either homozygotes or compound heterozygotes for RP1 mutations were ascertained (table 1), including 6 originating from France and Morocco,15 ,20 11 from Spain,17 ,21 4 from Saudi Arabia,18 3 from Pakistan11and 2 from the Netherlands.16 ,19 Patients who had cataract surgery or aphakic eyes were excluded from the study. For each patient, the refractive error (sphere, cylinder and axis), visual acuity and age were noted.
For refractive error comparisons with other forms of RP, we sorted out from the Montpellier's database 25 patients with autosomal dominant RP (adRP) due to RP1 mutations, 8 and 33 male patients with RP2 and RPGR mutations, respectively, 93 patients with autosomal recessive RP (arRP) without RP1 mutations and 198 patients with Usher syndrome types I or II (table 2).2
Besides standard ophthalmic examination to ascertain the presence of RP, the objective refraction was determined by an automatic refractometer (Nidek AR20) without cycloplegia. The best corrected visual acuity was assessed with the Snellen decimal chart and expressed as the logarithm of the minimum angle of resolution (logMAR). The SE was calculated using SE=sphere+0.5*cylinder. Emmetropia was defined when SE was between −0.5 and +0.5 D.
Median ages between groups were compared with the Wilcoxon test. For each series of patients, right eye (RE) and left eye (LE) were analysed separately. Because the distribution of refractive error significantly departed from normality in the majority of RP groups, all values are presented using median and IQR. The distribution of refractive errors in each group was presented using histograms and best fitted curve from kernel density estimation. While refractive errors displayed a skewed distribution, multiple comparisons between the refraction in the six RP groups were performed using the Kruskal–Wallis test. Pairwise comparisons were performed with Mann–Whitney test and corrected for multiple testing using the Bonferroni method. All analyses were performed using Statistical Analysis Software (SAS V.9.2; SAS Institute, Cary, North Carolina, USA).
The median of age of the patients was 35.0 years (IQR: 22–47 years). adRP patients (54.0 years, 37–63) with RP1 mutations were older than patients of all other RP groups (p<0.001) (table 2). In the group of arRP with RP1 mutations, most of the patients had myopia, with the majority of them showing peripapillary atrophy at funduscopy. There was no patient with severely abnormal corneal topography (keratoconus) or myopic staphyloma. The mean age of detection of myopia and RP was 6 (n=18, range 2–14) and 7 (n=21, range 1–14), respectively, indicating that both conditions occurred before adulthood. There was no difference between these ages.
We determined the SE in the 26 patients with RP1-associated arRP. As presented in table 3, the SE in this group was −4.0 (range −9.0; −2.13) dioptres (D) for RE and −3.88 (range −8.63; −1.88) D for LE. All patients had truncating mutations, either non-sense or frameshift mutations (table 1). We found that the level of the refractive error was independent of the amino acid position of the mutation (figure 1A).
In contrast to RP1-associated arRP patients, the 93 patients with arRP without RP1 mutations and the 198 patients with Usher syndrome who have typical signs of RP had a smaller myopic refractive error. Indeed, the median SE for arRP without RP1 mutations was −0.50 (range −1.88; +0.50) D for the RE and −0.75 (−2.00; +0.50) D for the LE, and for the Usher syndrome −0.50 (−2.25; +0.50) D for the RE and −0.38 (−2.00; +0.63) D for the LE. Table 4 shows p values for post-hoc pairwise comparisons corrected using the Bonferroni method. A significant difference in the refraction values was observed between RP1-associated arRP compared with arRP caused by mutations in other genes (p<0.0001) and Usher syndrome (p<0.0001).
To test whether the high myopic refractive error in case of RP1 mutation was dependent on the presence of a mutation in one or both alleles, we determined the refractive error in 25 adRP patients with a heterozygote mutation in RP1. We found that these patients had a moderate myopic refractive error with an SE at −0.13 (range −2.50; +0.88) D for the RE and at 0 (range −2.50; +0.13) D for the LE (table 3), not statistically different from the refraction found in arRP patients without RP1 mutations (p=0.97) or in Usher syndrome patients (p=0.81) (table 4), but significantly different from the refraction encountered in arRP patients with recessive RP1 mutations (p=0.00015) (table 4). This suggested that myopia in case of RP1 mutations is dependent on the presence of two mutated RP1 alleles, therefore occurring in case of presumably absent RP1 protein.
Since it is known that the patients with xlRP associated to either RPGR or RP2 mutations have often high myopic refractive errors, we also determined the refraction in xlRP patients from the Montpellier's clinical files and compared them with those of RP1-associated arRP patients. We found that xlRP patients with an RPGR mutation had an SE at −4.50 (range −8.50; −1.13) D for the RE and at −5.25 (range −7.88; −1.13) D for the LE (table 3). Similarly, RP patients with an RP2 mutation had an SE at −6.25 (range −8.56; −2.95) for the RE and at −6.88 (range −9.06; −2.44) D for the LE (table 3). When we compared the three groups, RP1-associated arRP, xlRP with RPGR or RP2 mutations, there were no significant differences between them; arRP RP1 vs xlRP RP2 p value for RE and LE=10.95 and 7.35, respectively, and arRP RP1 vs xlRP RPGR p value for RE and LE=9.30 and 14.6, respectively. This indicates that the patients in those three groups had similar levels of refractive error (table 4). As expected, myopic refractive error in xlRP with RPGR or RP2 mutations were, like for RP1, significantly higher than arRP patients without RP1 mutations and Usher syndrome patients (table 4). Astigmatism values for each group of patients are shown in table 5. In contrast to the SE, there were no differences between most groups except for adRP patients with RP1 mutations when compared with patients of other groups, and only in the RE (p=0.007), but not in the LE (p=0.386). This indicates that the astigmatism values were not related to differences seen in myopic refractive errors.
We then compared the distribution of the refractive errors in the six groups of patients (figure 1B). We observed that in the arRP without RP1 mutations, in Usher syndrome and in RP1-associated adRP, there was a small range of refractive errors with the IQR (Q25–Q75) being for the three groups between −2.50 and +0.88 D. In contrast, in RP1-associated arRP and in xlRP associated with RP2 or RPGR mutations, there was a larger range of refractive errors with the IQR being between −9.06 and −1.13 D for the three groups. We found that the highest percentage of emmetropic or hyperopic patients belonged to the groups of arRP without RP1 mutations (41% for RE, 37% for LE), of Usher syndrome (39% for RE, 40% for LE) and of RP1-associated adRP (48% for both RE and LE), whereas there were fewer emmetropic or hyperopic cases in the groups of patients with RP1-associated arRP (3.84% for both RE and LE), with RP2-associated (0% for both RE and LE) and RPGR-associated (12% for both RE and LE) xlRP (p=0.04). The shape of the curve for RP1-associated arRP and for RPGR-associated xlRP was similar. Indeed, besides some patients who were emmetropic, and with a moderate hyperopic or myopic refractive error, there were others in both groups with high myopic refractive error. This observation applies only partially to the RP2 group, in which there is an even distribution of myopic patients. Altogether, these findings indicate that an important proportion of the patients in these three groups had elevated myopic refractive error.
In this study, we show that the patients having arRP without RP1 mutations, Usher syndrome or adRP tend to have small refractive errors. Although there is a minor subset of patients with high myopic or hyperopic refractive error in the three groups, this is compatible with statistical probability of refractive errors found in the general population.22 Studies undergone before the molecular genetics era revealed that there was mild myopia in series of patients with arRP and adRP when compared with a population of 1056 normal men aged 17–27 in which the mean spherical refractive error was +1.007 D.4 ,5 ,23 However, in a more recent study involving 5018 eyes of a normal population with a mean age of 48 (range 22–84), the mean SE varied from −0.75 to −1.04 D, not different from the mean SE that we found in the three groups of patients having arRP without RP1 mutations, Usher syndrome or adRP, varying from −0.58 to −1.18 D depending on the eyes and groups.22 Therefore, the majority of patients from these three groups were not different from individuals of the general population.
In contrast to the group of arRP patients without RP1 mutations, we found that arRP patients with RP1 mutations have on average high myopic refractive error, not different from the myopia observed in xlRP caused by either RPGR or RP2 mutations.7 ,8 The xlRP patients reported in Sieving and Fishman5 could be subdivided in one group with moderate myopic refractive error (range +2.00 to −4.00) with a mode at −0.5 D and a second group with high myopic refractive error (range −4.00 to −16.00) with a mode at −7.50 D. This is similar to what we have found in our arRP group with RP1 mutations and in xlRP group caused by RPGR mutations. Myopia values in the arRP group with RP1 mutations were independent of age (data not shown). One can argue that the finding of high myopic refractive error in the RP1-associated arRP could be due to a myopia disease inherited as an autosomal dominant trait. This could be the case for a minority of the patients. However, it seems unlikely that autosomal dominant myopia disease would be responsible for high myopic refractive error in this group of patients, as it should be observed in the other arRP subgroups also.
To our knowledge, RP1 is the third RP gene, after RPGR and RP2, in which a relationship between non-syndromic RP and myopia is established. Yet, high myopic refractive error is also encountered in complete and incomplete forms of congenital stationary night blindness (CSNB), a non-progressing autosomal or X-linked recessive condition with mutations in genes expressed in the synapse of photoreceptors with bipolar cells.24 Myopia in CSNB is present in early life, usually found with an onset before 5 years of age. One can hypothesise that a degraded signal transmission in the retina from birth may favour the occurrence of myopia.
We noted that the level of the refractive error was independent of the position of the mutation. Yet, since all mutations are truncating, it is expected that they exert their pathological effect via a unique mechanism, which is the loss of function of the protein. Therefore, besides the mutations in RP1, there should be other factors influencing the degree of refractive error. It is striking that the median and the distribution of the myopia values are not very different in the three groups of RP patients, that is, with RPGR, RP2 and recessive RP1 mutations. It could be possible that a pathophysiological link exists between the three genes underlying myopia. RP1 is a photoreceptor-specific protein expressed in both rod and cone photoreceptors. It localises to the axoneme of the photoreceptor's outer segment, and hence plays an important role in the organisation of the outer segments. RP1, through two doublecortin domains located in tandem in its N-terminal part, directly interacts with the axonemal microtubules and also with the ciliary male germ cell-associated kinase (MAK), a regulator of the cilium length. The RP G-protein regulator encoded by RPGR is found in the transition zone of the cilium, where it interacts with RPGRIP1, which is bound to the microtubules.25 RP2 is also encountered in the transition zone of the cilium. RPGR and RP2 are important players in the trafficking machinery of the cilium. Together with ARL3 and PDE6 delta, they form a dynamic protein complex whose role is to cargo phototransduction proteins like catalytic subunits of the phosphodiesterase to the disc membrane of the photoreceptors by using the guiding structure of the cilium. Therefore, it seems plausible that the myopia observed in the three RP1, RP2 and RPGR subtypes of recessive RP is related to their function in the photoreceptor cilium. However, no RP or retinal dystrophies caused by mutations in other ciliary genes, including RPGRIP1, and genes for Joubert, Senior–Loken and Bardet–Biedl syndromes, are known to be specifically associated with myopia. Therefore, the pathophysiological link between the three ciliary proteins, RP1, RP2 and RPGR, and the occurrence of myopia awaits more detailed biochemical studies.
In conclusion, we have established that myopia is part of the characteristic features encountered in arRP due to RP1 mutations and that it is similar to myopia found in RP2 and RPGR xlRP. We suggest that in cases of arRP with high myopic refractive error, one should preferentially screen the RP1 gene.
We thank the patients and their families for participating in this study. The work was supported by Fédération des Aveugles et Handicapés Visuels de France, Information Recherche sur la Rétinite Pigmentaire, Retina France, SOS Retinite, UNADEV, Spanish National Organization for the Blind—ONCE, Spanish Foundation Fighting Blindness—FUNDALUCE and public funding (Health Research Fund (FIS: PI13-00226) from the Spanish Ministry of Health). MC is sponsored by the Miguel Servet Programme from the Instituto de Salud Carlos III (Spanish Ministry of Health); AAF is sponsored by CIBERER.
Contributors Conception and design, analysis and interpretation, writing the article, gene sequencing, final approval of the article, data collection, statistical expertise, obtaining funding, literature search, clinical data and patient examination: TC, BB, VD, AA-F, CA, RWJC, MC, JFH, LIvdB, BJK, SAR, NS, AL, IM, and CPH
Funding Fédération des Aveugles et Handicapés Visuels de France, Information Recherche sur la Rétinite Pigmentaire, Retina France, SOS Retinite, UNADEV, Spanish National Organization for the Blind—ONCE, Spanish Foundation Fighting Blindness—FUNDALUCE, Miguel Servet programme of the Instituto de Salud Carlos III (Spanish Ministry of Health), CIBERER.
Competing interests None declared.
Patient consent Obtained.
Ethics approval French Ministry of Health for biomedical research.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.