The Clinical and Mutational Spectrum of Bardet–Biedl Syndrome in Saudi Arabia

The retinal features of Bardet–Biedl syndrome (BBS) are insufficiently characterized in Arab populations. This retrospective study investigated the retinal features and genotypes of BBS in Saudi patients managed at a single tertiary eye care center. Data analysis of the identified 46 individuals from 31 families included visual acuity (VA), systemic manifestations, multimodal retinal imaging, electroretinography (ERG), family pedigrees, and genotypes. Patients were classified to have cone–rod, rod–cone, or generalized photoreceptor dystrophy based on the pattern of macular involvement on the retinal imaging. Results showed that nyctalopia and subnormal VA were the most common symptoms with 76% having VA ≤ 20/200 at the last visit (age: 5–35). Systemic features included obesity 91%, polydactyly 56.5%, and severe cognitive impairment 33%. The predominant retinal phenotype was cone–rod dystrophy 75%, 10% had rod–cone dystrophy and 15% had generalized photoreceptor dystrophy. ERGs were undetectable in 95% of patients. Among the 31 probands, 61% had biallelic variants in BBSome complex genes, 32% in chaperonin complex genes, and 6% had biallelic variants in ARL6; including six previously unreported variants. Interfamilial and intrafamilial variabilities were noted, without a clear genotype–phenotype correlation. Most BBS patients had advanced retinopathy and were legally blind by early adulthood, indicating a narrow therapeutic window for rescue strategies.


Introduction
Inherited retinal disorders (IRDs) comprise a heterogeneous group of diseases that manifest with or without systemic associations.Some IRDs are due to perturbations of the development and maintenance of the photoreceptor cilium and other primary cilia in the body.Bardet-Biedl syndrome (BBS) (MIM 209900) is an autosomal recessive ciliopathy and is the second most common form of syndromic retinopathy after Usher syndrome [1].The diagnosis of BBS is established if the following criteria are met: (1) the presence of four primary features or (2) the presence of three primary features with at least two secondary features [2].The primary features are retinal dystrophy, truncal obesity, postaxial polydactyly, hypogonadism, cognitive impairment, and renal anomalies.The secondary features are developmental delay, ataxia, poor coordination, short stature, speech disorder, behavioral abnormalities, strabismus, cataract, astigmatism, dental abnormalities, high arched palate, craniofacial dysmorphism, brachydactyly, syndactyly, lower limb spasticity, reproductive abnormalities, diabetes mellitus, cardiovascular anomalies, and hepatic fibrosis [2].
Previous studies described the clinical and molecular features of BBS in various endogamous and exogamous populations, but such data are sparse in the Arab population [1,9,10].The present study expands the spectrum of BBS mutations in the Saudi population, includes previously unreported variants, and describes the associated retinal phenotypes.

Methods
This is a retrospective study of 46 individuals, from 31 families, who harbored biallelic variants in known BBS genes and were reviewed at King Khaled Eye Specialist Hospital (KKESH), a tertiary eye care center in Riyadh, Saudi Arabia, from 2007 to 2019.Institutional Review Board (IRB)/Ethics Committee approval at KKESH was obtained (IRB: RP1933-R) and the study adhered to the tenets of the declaration of Helsinki.
Informed consent for clinical genetic testing was obtained from all participants.Data analyzed included: the earliest visual symptoms, age at onset of visual symptoms and last visit, sex, visual acuity at presentation and at last clinic visit, reported systemic features, and family history.All patients underwent full ophthalmic examination including Snellen VA, slit lamp biomicroscopy, and dilated fundoscopy.The retinal structure was evaluated with multimodal imaging: fundus images (Topcon, Tokyo, Japan; and Optos TM, Dunferline, Scotland, UK), widefield medium wavelength (532 nm) fundus autofluorescence (FAF, on the Optos machine), spectral-domain optical coherence tomography (OCT, Heidelberg Engineering, Heidelberg, Germany).Full-field electroretinography (ERG, Nicolet Biomedical Instruments, Madison, WI, USA; and Roland Consult, Brandenburg an der Havel, Germany) was performed using a protocol that was modified from the International Society for Clinical Electrophysiology of Vision Standards, as previously described [11].
Classification of FAF images, based on the patterns deduced from the macular AF changes, was performed jointly by two retina specialists (DM and RB) and was reviewed independently by a third retina specialist (SRN); in the case of discrepancy, RB reclassified the image(s).Since the ERGs were undetectable in 34 patients, and assuming that most patients with cone-rod dystrophy (CRD) would manifest with early severe maculopathy and foveal involvement, the retinal phenotype was classified as (1) rod-cone dystrophy (retinitis pigmentosa): if the VA at presentation was better than 20/200, with an annulus of increased AF at the macula and a sub-foveal EZ depicted on OCT (Figure 1(A1,A2,A3)); (2) CRD: if the VA at presentation was less than or equal to 20/200-used as a measure for foveal dysfunction, and the macular AF showed a pattern reminiscent of a bull's eye lesion with disrupted or hyporeflective ellipsoid zone (EZ) on OCT (Figure 1(B1,B2,B3)); (3) generalized photoreceptor dystrophy if the VA was less than 20/200, the AF did not show a distinct macular annulus of increased signal, and the OCT showed an absent or markedly disrupted EZ (Figure 1(C1,C2,C3)) [12,13].
Genetic testing was performed at two clinical laboratories using next-generation sequencing for the following retinal panels (Figure S1): autosomal recessive retinitis pigmentosa, BBS, cone-rod dystrophy (CRD), cone dystrophy (CD), and macular dystrophy as previously described [14].Classification of variants, based on the laboratory reports, was also scrutinized using Varsome 11.8 [15]; previously unreported variants were classified according to the ACMG guidelines [16].Genetic testing was performed at two clinical laboratories using next-generation sequencing for the following retinal panels (Figure S1): autosomal recessive retinitis pigmentosa, BBS, cone-rod dystrophy (CRD), cone dystrophy (CD), and macular dystrophy as previously described [14].Classification of variants, based on the laboratory reports, was also scrutinized using Varsome 11.8 [15]; previously unreported variants were classified according to the ACMG guidelines [16].

Patient Characteristics and Clinical Features
Seventy-one individuals with a clinical diagnosis of BBS were identified; twenty-five of them were not molecularly characterized and were excluded.Forty-six patients from thirty-one families were included.A summary of their demographic data is given in Table 1.The age range was 5-35 years (median: 19 years).Twenty-seven patients were males (59%).Poor night vision and subnormal visual acuity were the initial symptoms and noted during the first decade of life in 70% of the cases.Nystagmus was noted at presentation in seventeen patients (37%).Initial VA was 20/200 (1.00 logMAR) or less in the better seeing eye in 27/46 patients (58.7%), better than 20/200 in 12/46 (26.1%), and not measured in 7 patients who were able to fixate and follow but had limited cooperation.Of the 12 with initial VA better than 20/200, vision deteriorated to 20/200 or worse in five patients (41.6%; duration 1-10 years, mean 6.4).At the last clinic visit, the VA was 20/200 or worse in the better-seeing eye in 35/46 patients (76%) and better than 20/200 in 9/46 (19.6%); the remaining two patients were able to fixate and follow.The most frequent systemic features were obesity 42/46 (91.3%), polydactyly 26/46 (56.5%), and severe cognitive disability 15/46 (32.6%).

Molecular Genetics
The probands from 31 families were genotyped (Figure S6).Twenty patients were simplex cases and underwent genotyping and phenotyping.For families with multiple affected members, genotyping was carried out only for the probands, but the clinical data were analyzed for all the affected individuals who presented to the IRD clinic.Nine probands (29%) had variants in BBS4, six (19.4%) in MKKS, four (13%) in BBS1, and three (9.7%) in BBS5.Eight probands had variants in one of the following genes: BBS9, BBS10, BBS12, ARL6 (6.5% each), and one proband had variants in BBS2 (3.2%) (Table 3).Nineteen probands (61%) harbored a mutation in one of BBSome complex genes, ten (32%) harbored a mutation in one of chaperonin complex genes (Table 3).Two probands (6%) were homozygous for variants in ARL6.
Because most patients had undetectable ERGs, the patients in this cohort were classified according to the VA, the macular AF and OCT findings as described in the Methods section.Of the 40 patients who had AF and OCT images, thirty (75%) were classified to have cone-rod dystrophy, six (15%) had generalized photoreceptor dystrophy, and four (10%) had rod-cone dystrophy (Table 2).

Molecular Genetics
The probands from 31 families were genotyped (Figure S6).Twenty patients were simplex cases and underwent genotyping and phenotyping.For families with multiple affected members, genotyping was carried out only for the probands, but the clinical data were analyzed for all the affected individuals who presented to the IRD clinic.Nine probands (29%) had variants in BBS4, six (19.4%) in MKKS, four (13%) in BBS1, and three (9.7%) in BBS5.Eight probands had variants in one of the following genes: BBS9, BBS10, BBS12, ARL6 (6.5% each), and one proband had variants in BBS2 (3.2%) (Table 3).Nineteen probands (61%) harbored a mutation in one of BBSome complex genes, ten (32%) harbored a mutation in one of chaperonin complex genes (Table 3).Two probands (6%) were homozygous for variants in ARL6.
Family pedigrees were available for 23 of the 31 probands (Figure S6).Twenty-two probands (71%) came from consanguineous families, and harbored homozygous variants in one of the BBS genes, except for family 23 where the proband was heterozygous for two variants in BBS4 (Table 3).Six probands came from non-consanguineous families, four harbored homozygous variants in BBS1, BBS10, and MKKS; and two harbored two heterozygous variants in BBS2 and BBS1.Three probands came from non-consanguineous families, but the parents originated from the same tribe, and they harbored homozygous variants in BBS1, BBS4, and BBS12.

Genotype-Phenotype Correlation
Clinical classification of retinopathy showed a continuum where the predominantly affected photoreceptor cell type was not always determined.All but four patients had syndromic features (Table 1).The four probands with apparently non-syndromic CRD or generalized photoreceptor degeneration had biallelic variants in BBS4 (proband 6), ARL6 (proband 11) and BBS6 (probands 12 and 25) (Tables 1 and 3).
The proband from family 13 was homozygous for a variant in MKKS: c.116C>T, p.(Pro39Leu); while one sibling (13A) manifested with rod-cone dystrophy, the other two (13B and 13C) had CRD.Family 30 harbored the same variant in MKKS, with one member (30B) manifesting with rod-cone dystrophy, while the other sibling (30A) manifested with CRD.
There was no correlation between the genotype and the severity of central visual loss in this study.Nine patients had VA > 20/200 in the better seeing eye on the last clinical examination.The age range was 9-20 years (median: 16 years) (Tables 2 and 3).These patients harbored biallelic variants in genes encoding components of the BBSome, as well as the chaperonin complex: BBS1, BBS4, BBS5, BBS9, BBS12, MKKS.Twenty-seven patients had visual acuity ≤ 20/200 at presentation (age 6-32 years, median: 20 years).Those patients harbored biallelic variants in the same genes found in the former group in addition to ARL6, BBS10, BBS2.

Discussion
In this study, the retinal and main systemic manifestations of BBS were described in a cohort of Saudi patients, managed at the country's largest tertiary eye care center.We also identified six previously unreported variants in four BBS genes.
Most patients in this study had advanced retinal degeneration such that phenotypic classification to have either rod-cone dystrophy or CRD was not feasible based on ERG, as previously reported [26,39].Other parameters that aid in phenotyping are clinical history and retinal imaging.Although the first two symptoms commonly reported were nyctalopia and reduced visual acuity, due to foveal involvement, the latter suggests that the foveal cones are particularly vulnerable to dysfunctional BBS proteins.Moreover, information can be gained from retinal imaging.For example, the distribution of macular AF changes, such as the parafoveal ring of increased signal with normal signal in the middle, as seen in rod-cone dystrophy, or altered central signal as seen in cases of cone-rod dystrophy is better defined than mid-peripheral AF where the signal can be either indistinct (featureless) or lost, due to deep retinal changes or intra-retinal pigment migration and loss of the retinal pigment epithelium, respectively [12,13].Both mid-peripheral changes can occur in rodcone dystrophy or CRD.Macular OCT is also useful in assessing the integrity of the foveal ellipsoid zone and assessing the degree of loss of lamination, due to retinal remodeling [40].The classifications proposed in this study suggest that only a minority of patients had typical retinitis pigmentosa.This differs from other reports of large cohorts, which showed, based on ERG recordings, a rod-cone dystrophy pattern [3].This difference is due to our reliance on retinal imaging, highlighting a discrepancy between retinal features and the retinal mass responses on ERG.
Four patients in this study, carrying four different homozygous variants in BBS4, ARL6, and MKKS genes, were found to have no obvious syndromic features.Previously, variants in C8orf37 [8], ARL6 [41,42], and BBS8 [43] were associated with non-syndromic retinopathy [44].The absence of syndromic features could be due to the presence of retinaspecific isoforms of the same gene, or the involvement of amino acids that are crucial for the photoreceptor function but have no clinically significant impact on the cilia in other organs.
The most common genes in this study were BBS4 (29%) and MKKS (19.4%).This finding differs from other national studies from a general hospital setting where ARL6, BBS1, and BBS2 were the most frequently mutated genes [10,22].Twenty-five patients in this study were excluded as their genotypes were not available, and this might have contributed to this difference.Additionally, patients with significant visual loss, particularly those with non-syndromic retinopathy or only subtle features, are usually diagnosed with BBS by ophthalmologists, whilst those with prominent systemic features present first to general hospitals.Internationally, the most frequently mutated BBS genes were BBS1 and BBS10 [4,38,45].
The high prevalence of homozygous variants in this study offers an opportunity to assess the effect of these variants on the retina and aids in the diagnosis of patients who may harbor one of these variants in trans with a novel variant.Intrafamilial variability of systemic manifestations of BBS was documented in the literature; however, little is known about variability of the retinal phenotype within the same family [46,47].As previously reported, there was no clear genotype-phenotype correlation in our cohort [45,[48][49][50].Additionally, we identified a recurrent variant in BBS4: c.157-2A>G causing both CRD and rod-cone dystrophy.Similarly, a recurrent variant in MKKS: c.116C>T, p.(Pro39Leu) caused CRD and rod-cone dystrophy in different individuals.As observed by others, variants in the same BBS gene have been reported to give rise to either CRD or retinitis pigmentosa [26,51].
Inherited retinal disorders were recently ranked at the top of the causes of blind registration in the working age group in developed countries [52].Given that the median age in this study was 19 years, and 76% of the patients were legally blind at the last clinic visit, it is reasonable to conclude that BBS would have a stronger socio-economic impact compared to non-syndromic IRDs, as patients are also affected by other comorbidities.Future therapeutic trials for patients with BBS should consider targeting the systemic manifestations as well as multiple cell types in the retina since it has been suggested that BBS proteins are also expressed in other retinal cell types [53,54].The early presentation and rapid progression of visual loss in BBS patients impose a narrow window of opportunity for novel therapeutic interventions such as gene rescue; therefore, other approaches such as optogenetics and visual rehabilitation would be more suitable for advanced retinopathy.Premarital screening in highly consanguineous populations, genetic counselling, and a multidisciplinary approach remain the standard of care for BBS patients.
In conclusion, this study has the largest Middle Eastern cohort to depict BBS as a severe form of retinopathy in our population.Additionally, the study added six previously unreported variants to the genetic spectrum of BBS.Due to the retrospective nature of the current study, the depth of systemic phenotyping was limited, which may have led to underreporting of BBS systemic manifestations.

Figure 1 .
Figure 1.Ultra-widefield color fundus images, fundus autofluorescence (FAF) and optical coherence tomography (OCT) of the three retinal phenotypes: rod-cone dystrophy (RCD), cone-rod dystrophy (CRD) and generalized photoreceptor involvement.(A) (A1,A2,A3) An example of RCD; (A1) color photo showing widespread retinal alterations, vascular attenuation and multiple hypopigmented small patches in the mid-periphery (black arrow); (A2) FAF revealing an annulus of increased AF at the macula (white arrow); the hypopigmented patches in (A1) co-localize with hypo-autofluorescent patches in the mid-periphery indicating retinal atrophy; (A3) OCT depicting relatively spared subfoveal ellipsoid zone (EZ) (yellow arrow).(B) (B1,B2,B3) An example of CRD; (B1) color image showing widespread retinal alterations with widespread patchy hypopigmentation (black arrow), note the relatively milder vascular attenuation compared to (A), and dull foveal reflex; (B2) FAF shows a patch of increased signal at the macula surrounded by reduced signal (white arrow); there is widespread patchy hypo-AF in the mid-periphery; (B3) OCT showing sub-foveal hyporeflective EZ which tapers abruptly at the edges of the fovea with loss of the outer nuclear layer (yellow arrow).(C) (C1,C2,C3) An example of generalized photoreceptor involvement; (C1) color image showing bone spicule-like pigmentation extending from the vascular arcades to the periphery (black arrow), vascular attenuation and macular atrophy; (C2) FAF revealed diffuse hypo-AF signal in the mid-peripheral retina and macula and a distinct macular annulus of retained signal surrounding a patch of signal loss (white arrow); (C3) OCT depicted an absent EZ (yellow arrow), epiretinal membrane and severe laminar disorganization.

Figure 1 .
Figure 1.Ultra-widefield color fundus images, fundus autofluorescence (FAF) and optical coherence tomography (OCT) of the three retinal phenotypes: rod-cone dystrophy (RCD), cone-rod dystrophy (CRD) and generalized photoreceptor involvement.(A) (A1,A2,A3) An example of RCD; (A1) color photo showing widespread retinal alterations, vascular attenuation and multiple hypopigmented small patches in the mid-periphery (black arrow); (A2) FAF revealing an annulus of increased AF at the macula (white arrow); the hypopigmented patches in (A1) co-localize with hypo-autofluorescent patches in the mid-periphery indicating retinal atrophy; (A3) OCT depicting relatively spared subfoveal ellipsoid zone (EZ) (yellow arrow).(B) (B1,B2,B3) An example of CRD; (B1) color image showing widespread retinal alterations with widespread patchy hypopigmentation (black arrow), note the relatively milder vascular attenuation compared to (A), and dull foveal reflex; (B2) FAF shows a patch of increased signal at the macula surrounded by reduced signal (white arrow); there is widespread patchy hypo-AF in the mid-periphery; (B3) OCT showing sub-foveal hyporeflective EZ which tapers abruptly at the edges of the fovea with loss of the outer nuclear layer (yellow arrow).(C) (C1,C2,C3) An example of generalized photoreceptor involvement; (C1) color image showing bone spicule-like pigmentation extending from the vascular arcades to the periphery (black arrow), vascular attenuation and macular atrophy; (C2) FAF revealed diffuse hypo-AF signal in the mid-peripheral retina and macula and a distinct macular annulus of retained signal surrounding a patch of signal loss (white arrow); (C3) OCT depicted an absent EZ (yellow arrow), epiretinal membrane and severe laminar disorganization.

Figure 2 .
Figure 2. Ultra-widefield color fundus images and fundus autofluorescence (FAF) of three different patients with distinct features.(A) Patient 28B has Coats'-like picture; (A1) color image showing macular and peripheral exudation and telangiectatic blood vessels (white arrow); (A2) FAF revealed patches of decreased AF inferior to the macula and in the midperiphery with loss of AF at the macular center, a ring of loss of AF (cryotherapy mark) is noted supero-temporally (black arrow).(B) Patient 30A has cone-rod dystrophy and congenital hypertrophy of the retinal pigment epithelium (CHRPE) lesion temporally; (B1) color image showing vascular attenuation, and a bull's eye lesion at the macula; a round pigmented lesion comprising lacunae (CHRPE, white arrow); (B2) FAF revealed round area of decreased AF temporally (black arrow), numerous small patches of decreased AF at the midperiphery and anterior to the arcades with a central patch of increased AF at the macular center, surrounded by an annulus of decreased AF. (C) Patient 18B has nummular pigmentation; (C1) color image showing retinal atrophy, severe vascular attenuation, bone spicules and nummular pigmentation in the nasal midperiphery (white arrow), a CHRPE lesion temporally (yellow arrow) and a bull's eye lesion at the macula.(C2) FAF revealed patches and nummular dots of decreased AF at the midperiphery (black arrow), round area of decreased AF temporally (yellow arrow) and adjacent to the arcades with a central patch of increased AF at the macular center, surrounded by an annulus of decreased AF

Figure 2 .
Figure 2. Ultra-widefield color fundus images and fundus autofluorescence (FAF) of three different patients with distinct features.(A) Patient 28B has Coats'-like picture; (A1) color image showing macular and peripheral exudation and telangiectatic blood vessels (white arrow); (A2) FAF revealed patches of decreased AF inferior to the macula and in the midperiphery with loss of AF at the macular center, a ring of loss of AF (cryotherapy mark) is noted supero-temporally (black arrow).(B) Patient 30A has cone-rod dystrophy and congenital hypertrophy of the retinal pigment epithelium (CHRPE) lesion temporally; (B1) color image showing vascular attenuation, and a bull's eye lesion at the macula; a round pigmented lesion comprising lacunae (CHRPE, white arrow); (B2) FAF revealed round area of decreased AF temporally (black arrow), numerous small patches of decreased AF at the midperiphery and anterior to the arcades with a central patch of increased AF at the macular center, surrounded by an annulus of decreased AF. (C) Patient 18B has nummular pigmentation; (C1) color image showing retinal atrophy, severe vascular attenuation, bone spicules and nummular pigmentation in the nasal midperiphery (white arrow), a CHRPE lesion temporally (yellow arrow) and a bull's eye lesion at the macula.(C2) FAF revealed patches and nummular dots of decreased AF at the midperiphery (black arrow), round area of decreased AF temporally (yellow arrow) and adjacent to the arcades with a central patch of increased AF at the macular center, surrounded by an annulus of decreased AF.
c.966dupT: p.(Ala323Cysfs*57) [29] c.966dupT: p.(Ala323Cysfs*57) Pathogenic : gene panels of the two clinically accredited laboratory centers, Figure S2: fundus images, Figure S3: fundus autofluorescence images, Figure S4: (a) OCT images, (b) fundus images; Figure S5: ERG data, Figure S6: pedigree of the probands.Author Contributions: Conception and design: D.M. and R.B.-A.; data collection: D.M.; data analysis and interpretation: D.M., R.B.-A.and S.R.N.; manuscript writing and revision: D.M., R.B.-A.and S.R.N.All authors have read and agreed to the published version of the manuscript.Funding: This research received no external funding.Institutional Review Board Statement: This study was conducted in accordance with the Declaration of Helsinki and approved by Institutional Review Board (IRB)/Ethics Committee approval at KKESH (IRB no.: RP1933-R approval date 23 April 2019).Informed Consent Statement: Written informed consent has been obtained from the all patients.

Table 1 .
Demographic data and clinical features of the patients.Abbreviations: M: male, F: female.

Table 3 .
Molecular data of the 31 probands (novel mutations identified in this study are marked with an asterisk [*]).