Molecular Vision 2022; 28:300-316
<http://www.molvis.org/molvis/v28/300>
Received 21 September 2021 | Accepted 28 September 2022 | Published 01 October 2022
Molecular Vision 2022; 28:300-316
<http://www.molvis.org/molvis/v28/300>
Received 21 September 2021 |
Accepted 28 September 2022 |
Published 01 October 2022
Ulrika Kjellström, Sten Andréasson
Lund University, Skane University Hospital, Department of Clinical Sciences Lund, Ophthalmology, Lund, Sweden
Correspondence to: Ulrika Kjellström, Ögonkliniken Skånes Universitetssjukhus Lund, Kioskg 1, S 221 85 Lund, Sweden, Phone: +46-70-566 52 54, FAX: +46-46-13 90 45, email: ulrika.kjellstrom@med.lu.se
Purpose: To investigate whether the reduced retinal function and morphological retinal changes previously demonstrated in ABCA4 carriers had remained stationary or had deteriorated over time at 5-year follow-up to further explore if carriers of an autosomal recessive trait also express a weak phenotype, although this is not expected for an autosomal recessive disorder.
Methods: Thirteen ABCA4 carriers from a previous study that included parents to patients with well known genetically verified ABCA4-associated retinal degenerations were reexamined 5 years after the initial examination. As novel genes and new variants in already established genes are continuously reported, all subjects underwent renewed genetic testing with a next-generation sequencing (NGS) panel that included 288 genes associated with retinal dystrophies and an analysis of deep intronic mutations and copy number variations in the ABCA4 gene. Moreover, to evaluate any changes in retinal function and/or structure over time, clinical reassessment with Goldmann perimetry, visual acuity testing, fundus photography, fundus autofluorescence (FAF) imaging, optical coherence tomography (OCT), full-field electroretinography (ffERG), and multifocal ERG (mfERG) were performed 5 years after the initial investigation. The values of the ffERG parameters were compared between the two time points (the measurements obtained in the initial study versus the measurements at 5-year follow-up) and with the controls. The mfERG results of the carriers were compared with those of the controls.
Results: The renewed genetic testing confirmed the previously established ABCA4 mutations but also revealed the hypomorph ABCA4 variant c.5603A>T in five ABCA4 carriers. In three of them, the variant was found to be associated with known disease-causing alleles that always carry the c.5603A>T in cis. According to recent publications, the subjects could still be considered ABCA4 carriers because both variants are on the same allele. In the remaining two subjects, c.5603A>T could be in trans with the previously known ABCA4 variant, and the subjects were therefore excluded from the study since they could no longer be considered as carriers only. Statistical comparison of ffERG parameters showed significant reduction of the isolated rod, -as well as the combined rod-cone amplitudes over the five years of follow-up, but not compared with the controls. Concerning macular function, mfERG amplitudes were reduced for all rings in the carriers compared with the controls. Fundus photographs demonstrated morphological changes in 64% of the carriers, and 36% of them had further changes at follow-up. FAF images showed alterations in 55% of the carriers, with increased changes in 36% of them. Abnormalities on OCT were observed in 82% of the carriers, of whom 9% had newly found abnormalities at follow-up.
Conclusions: At 5-year follow-up, the ABCA4 carriers, who previously demonstrated reduced macular function, presented with deterioration of general retinal function, including reduced isolated rod and mixed rod-cone ffERG responses combined with a slight increase in morphological changes in some subjects. This indicates that carriership of at least some ABCA4 variants may cause a condition similar to a subgroup of dry age-related macular degeneration (AMD). In the long run, this might be of importance concerning the possibilities to also treat this subgroup of AMD patients with future gene-based and pharmacological drugs targeting ABCA4-associated disorders.
Compound heterozygous or homozygous mutations in the ABCA4 gene (Gene ID: 24, OMIM 601691) are known to cause a spectrum of ABCA4-associated retinal degenerations from the mildest form; Stargardt disease (STGD), mostly reducing macular function [1-5], to the more severe entities, including cone-rod dystrophy (CRD) [6-9] and autosomal recessive retinitis pigmentosa (arRP) [3,7,10], which lead to a more widespread reduction of retinal function. STGD is usually associated with typical fundus changes, including deep yellow-orange flecks in the posterior pole and symmetric and bilateral atrophies of the retinal pigment epithelium (RPE) in the macular region [1,2,4]. In CRD, fundus changes are less typical, with unspecific pigmentations and atrophies in the macula and midperiphery. Very severe cases can demonstrate pale optic discs and attenuated vessels [3,9,11]. Typically, patients with retinitis pigmentosa (RP) have peripheral bone corpuscle pigmentations, pale optic discs, and attenuated retinal vessels [3,7,10], but in ABCA4-associated arRP, retinal pigmentations can be less specific.
The ABCA4 gene is located on chromosome 1 (OMIM; 1p22.1). It encodes the ABCA4 protein, a membrane-associated protein in the rim of the photoreceptor discs [12-14]. ABCA4 is a transport protein belonging to the ATP-binding cassette (ABC) superfamily. It plays an important role in photoreceptors by removing the toxic retinoid compounds, N-retinylidene-phosphatidylethanolamine and phosphatidylethanolamine, from the inside of the outer segment discs to the cytosol during the phototransduction cascade [15-17]. Defective or missing ABCA4 protein results in a build-up of lipofuscine fluorophores such as N-retinylidene-N-retinylethanolamine in RPE cells [17-19], eventually leading to RPE cell death and probable secondary loss of photoreceptors [17,20]. Some surveys have also suggested that photoreceptor cell death is the primary process [21-25].
Previous studies have found indications of a potential role for heterozygous ABCA4 mutations in age-related macular degeneration (AMD) [4,26-29], although carriers of recessive traits are not supposed to express the phenotype. The carriership of pathological ABCA4 mutations is quite common, with an incidence of approximately 1 in 20 persons, in different populations [30,31]. Therefore, the role of the ABCA4 gene in AMD could be of interest to explain the cause of dry AMD, at least, in some subgroups of this widely prevalent disorder. Moreover, if an association with the ABCA4 gene can be confirmed, novel future gene-based and pharmacological treatments especially developed for ABCA4-associated disorders might also be of interest for this subgroup of patients with dry AMD who lack treatment options today.
The present study was a 5-year follow-up of ABCA4 carriers who, in the initial survey [27], presented with reduced macular function measured with multifocal electroretinography (mfERG) and, in some cases, morphological retinal changes on optical coherence tomography (OCT), fundus autofluorescence (FAF), and fundus photographs, although they carry only one pathological ABCA4 variant. These subjects were reexamined to evaluate whether the changes in retinal function and morphology were stationary or progressive to further explore the possibility of an association with AMD.
In this survey, ABCA4 carriers from a previous study [27] were reinvestigated 5 years after the initial examination. They were originally recruited as the parents to young patients with ABCA4-associated retinal dystrophies carrying two documented pathogenic ABCA4 variants that were followed up regularly in our department during the previous study period. None of the ABCA4 carriers had been examined in our clinic before the original study started, and only two of them, subjects 1 and 14, had had any subjective problems with their eyes or vision previously. Subject 1 underwent surgery for cataract in her right eye (in another hospital) and still had a mild cortical cataract in her left eye. Patient 14 had not had his right eye checked, although he had noticed some blurriness. Thus, the inclusion criterion for the subjects was the presence of a pathogenic ABCA4 variant on one allele. The exclusion criteria were the presence of pathogenic ABCA4 variants on both alleles, previous retinal surgery or serious eye trauma, and previous diagnosis of chronic eye disorders.
The initial study [27] originally recruited 18 Swedish parents. Sixteen of them were found to be carriers of a single heterozygous pathological ABCA4 variant and were thus included in the study, while two of them were compound heterozygous for two pathological ABCA4 mutations and were thus excluded. In the initial study, the ABCA4 carriers demonstrated reduced macular function and morphological changes in the posterior pole. In the present survey, the ABCA4 carriers were reexamined to evaluate whether the reduced retinal function and structural changes had been stationary over 5 years or if further deterioration could be detected. Moreover, because novel genes and new variants in already established genes are continuously reported, renewed genetic testing was also performed. Unfortunately, not all the original 16 ABCA4 carriers could participate in the 5-year follow-up survey. Two of them could not be reached, and a third subject declined to participate because of health problems in a close relative. Thus, 13 ABCA4 carriers aged 49–66 years (mean, 58 years) at the time of the reassessment were included. The subjects' demographic, genetic, and clinical data are described in Table 1.
In the present study, two control groups were included, one for comparison of mfERG parameters, consisting of 15 healthy subjects aged 49–64 years (mean, 55 years), and another for comparison of the full-field ERG (ffERG) parameters, including 20 healthy subjects aged 45–73 years (mean, 59 years). The inclusion criteria for the control subjects were the absence of retinal disorders or former retinal surgery, no history of serious eye trauma, and a best-corrected visual acuity (BCVA) of ≥0.9 on the letter chart. The exclusion criteria included a diagnosis of a chronic eye disorder, previous eye trauma or retinal surgery, and a BCVA of ≤0.8 on the letter chart. The study was conducted in accordance with the tenets of the Declaration of Helsinki and was approved by the Ethical Committee for Medical Research at Lund University. All subjects received a thorough written and oral explanation of the nature and possible consequences of the study and thereafter provided their written consent to participate.
BCVA was tested monocularly at 5 m on a letter chart, and Goldmann perimetry was performed, likewise monocularly, to map the visual fields using standardized objects V4e, I4e, 03e, and 02e. In eight subjects, the 01e object was also used. Fundus photographs, including color photographs, red-free photographs, and FAF images, were captured with a Topcon TRC 50DX fundus camera (Topcon, Inc., Oakland, NJ). The field of view for the FAF images was 50° × 50°, and Spaide filters with an excitation wavelength of 530–580 nm and barrier filters with a wavelength of 600–720 nm were used. In three subjects, careful slit-lamp and fundus examinations were performed.
In the initial study [27], DNA was extracted from a venous blood sample drawn from the precubital vein in all subjects. The sequence analysis of the entire coding region of the ABCA4 gene was performed at Asper Biotech (Asper Ophthalmics, Tartu, Estonia). Positive results were verified by Sanger sequencing. On follow-up, DNA was reanalyzed with an eye disease next-generation sequencing (NGS) panel that included 288 genes, with the addition of analyses for deep intronic mutations, copy number variations (CNVs), and hypomorph alleles in the ABCA4 gene.
ffERGs were recorded with an Espion E2 analysis system (Diagnosys, Lowell, MA) in accordance with the standardized guidelines for clinical ERG recommended by the International Society for Clinical Electrophysiology of Vision (ISCEV) [32,33], with the exception that the dark-adapted 10 ERG result was not recorded and that the 30-Hz flicker recordings were obtained during both dark- and light-adapted conditions, as our laboratory has used those settings for many years. Measurements were obtained after 40 min of dark adaptation with a bipolar Burian Allen corneal contact lens electrode. The ground electrode was placed on the forehead. The pupils were maximally dilated using cyclopentolate 1% and 10% phenylephrine hydrochloride, and the cornea was anesthetized with oxybuprocaine 0.4%. To ensure reproducibility, stimulations were repeated for each stimulus intensity until two sequential identical curves were registered.
Multifocal ERGs (mfERGs) were registered with a Visual Evoked Response Imaging System (VERIS Science 6; EDI, San Mateo, CA), with settings adhering to the ISCEV standards [32,33]. A Burian Allen bipolar ERG contact lens electrode was placed on the anesthetized (oxybuprocaine 0.4%) cornea, with the ground electrode on the forehead. The pupils were maximally dilated with cyclopentolate 1% and 10% phenylephrine. Fixation was monitored using an infrared eye camera mounted in the equipment. The stimulus matrix consisted of 103 hexagonal elements, scaled with eccentricity to elicit approximately equal amplitude responses at all locations. Each hexagon independently alternated between black and white according to a pseudorandom binary m-sequence at 75 Hz. The first-order component of the mfERG was analyzed for amplitudes (A) and implicit times (ITs) of P1 (first positive peak) within the six concentric rings (A 1–5 and IT 1–5) around the fovea. Ring 1, the innermost ring, included the summed responses from the central hexagon and the surrounding first ring of stimulus elements.
OCT images were captured, and the macular thickness was measured with a Topcon 3D OCT-1000 (Topcon, Inc., Paramus, NJ), as previously described [11,27].
Statistical analyses were performed using SPSS 26.0 (IBM SPSS Statistics, IBM Corporation, Chicago, IL). Statistical significance was defined as a p value <0.05. Owing to the small sample size, non-parametric statistical tests were used. To compare the ffERG results at the first investigation and those at 5-year follow-up, the Wilcoxon signed-rank test was used. For comparison of the values of the ffERG and mfERG parameters between the subjects and control groups, the Mann–Whitney U test was used. The limits of normality of the ERG parameters were defined by the mean ± two standard deviations (2SD). The lower limits of normality were used for amplitudes; and the upper limits of normality, for ITs.
The updated genetic analysis using NGS confirmed the ABCA4 mutations described previously [27] (Table 1) and revealed the hypomorphic ABCA4 variant c.5603 A>T in five subjects (subjects 4, 5, 8, 10, and 11). c.5603 A>T has been considered benign owing to its high minor allele frequency (approximately 7%; 34) in the general population but was recently confirmed to be associated with late-onset ABCA4-associated retinal degenerations [34,35]. In three subjects (subjects 4, 5, and 11), the c.5603A>T variant was found in association with the pathogenic alleles, c.5461–10T>C, c.319 C>T, and c.4469 G>A, which are known to always carry c.5603A>T in cis [34,35]. Hence, the subjects could still be considered ABCA4 carriers. The remaining two subjects (subjects 8 and 10) were carrying mutations (c.768 G>T and c.1804 C>T) that can be in trans with c.5603A>T [34,35]. Accordingly, the subjects might be compound heterozygous and were therefore excluded from the study. No deep intronic mutations or CNVs were found. Neither did any of the subjects carry pathogenic variants in other genes associated with retinal dystrophies.
The genetic reanalysis using NGS revealed that c.5603 A>T was the most common mutation, followed by c.768 G>T (four cases) and then c.2894 A>G (three cases).
Visual acuity and visual fields-- Visual acuity data, both at the initial and follow-up examinations, are presented in Table 1. Of the 11 ABCA4 carriers, four (36%) had a BCVA of 1.0 in both eyes, and two (18%) had a BCVA of <1.0 bilaterally. The BVCA had deteriorated at the time of follow-up in five (45%) of the 11 ABCA4 carriers but improved in two (18%). Only two subjects (1 LE and 14 RE) had noticed reduced vision in daily life. The reduced visual acuity was caused by moderate cortical cataract in Subject 1 and scarring in the macula in Subject 14 (Figure 1).
All subjects completed Goldmann visual field testing in both eyes. Short descriptions of the visual field results are presented in Table 2. In four (36%; subjects 1, 2, 5, and 7) of the 11 ABCA4 carriers, the result had deteriorated slightly at follow-up compared with the initial examination (Table 2). On the other hand, two (18%) of the 11 ABCA4 carriers had slightly wider isopters for 03e and 02e on follow-up.
Retinal structure-- The fundus photographs and examination results showed morphological retinal alterations in seven (64%) of the 11 ABCA4 carriers. Descriptions of the retinal morphologies are presented in Table 3. The most common findings were subtle pigmentary changes in the macula (Figure 2), but in one subject (Subject 1), drusen were scattered outside and just inside the vascular arcades (Figure 3A, B). In four (36%; subjects 1, 2, 6, and 12) of the 11 ABCA4 carriers, a slight aggravation of retinal changes was found at follow-up, whereas no remarkable features were found in the fundus photos of subjects 3, 4, 11, and 13 (36%). In subject 8, numerous orange-yellow retinal flecks (Figure 3C, D), which are quite typical for STGD, were found in the posterior pole both on the initial and follow-up examinations. These could be explained by the results of the renewed DNA analysis revealing the hypomorph variant c.5603A>T in addition to the previously known c.768 G>T variant. Thus, subject 8 now has both a phenotype and a genotype that agrees with late-onset STGD [34,35].
FAF images were captured in all subjects and showed alterations in six (55%) of the 11 ABCA4 carriers. The changes were subtle, with small foci of increased and/or reduced autofluorescence in the posterior pole (Table 3 and Figure 2). The FAF images of subject 8 showed more widespread changes with multiple pisiform foci of high and low autofluorescence (AF) scattered in the entire posterior pole and outside the vascular arcades (Figure 3), confirming the late-onset STGD phenotype in line with genetics. Four (subjects 5, 7, 12, and 14) of the 11 (36%) ABCA4 carriers had developed slightly more FAF changes in one eye at the time of follow-up (Figure 2).
OCT images were captured in both eyes of all ABCA4 carriers (22 eyes). Nine eyes showed normal layering and overall morphology on OCT b-scans, whereas 12 eyes demonstrated subtle attenuation of the ellipsoid zone. The right eye of Subject 14 showed more widespread morphological changes on OCT b-scans (Figure 1). Descriptions of the OCT findings are shown in Table 4 and examples of OCT images can be viewed in Figure 1 and Figure 4. Two carriers (subjects 4 and 11) had normal OCT images in both eyes, whereas four subjects (subjects 1, 6, 13, and 14) demonstrated morphological changes on OCT bilaterally. At follow-up, nine (82%) of the 11 ABCA4 carriers showed some degree of abnormality on OCT, and one carrier (6 LE) had developed new subtle alterations in an eye that was normal at the initial examination.
General retinal function-- All ABCA4 carriers completed ffERG assessments using the same equipment both at the initial and follow-up examinations. The statistical comparison of the ffERG parameters for the right eye was made between the two time points (Wilcoxon signed-rank test) and showed significantly reduced amplitudes for the isolated dark-adapted rod response (p=0.013) and the b-wave amplitudes of the dark-adapted combined rod and cone response (p=0.013) at follow-up. The a-wave amplitudes for the combined rod and cone responses were also reduced with borderline significance (p=0.050). Box plots illustrating the results are shown in Figure 5. The amplitudes for the isolated 30-Hz flicker cone responses were not significantly altered. Table 5 shows the mean values of the ffERG parameters for the subjects at the initial and follow-up examinations and the p values for the Wilcoxon signed-rank test results.
The comparison of right-eye ffERG parameters between the ABCA4 carriers and the controls (Mann–Whitney U test) did not show any significant differences. Nevertheless, five (45%; subjects 1, 2, 5, 6, and 12) of the eleven ABCA4 carriers had at least one significantly reduced ffERG parameter (Table 6) compared with the controls.
mfERGs were recorded in all subjects both at the initial and follow-up examinations. Unfortunately, the mfERG VERIS equipment was updated during the 5-year follow-up, making the comparison between the two time points ambiguous. Therefore, the statistical comparison between the two time points was omitted, and the values of the mfERG parameters were only compared between the ABCA4 carriers and the control group (Mann–Whitney U test), demonstrating reduced P1 amplitudes for all rings of the mfERG (A1–A5: A1, p=0.024; A2, p=0.017; A3, p=0.002; A4, p=0.001; and A5, p=0.006) and delayed P1 ITs for IT4 and IT5 (IT4: p=0.039; IT5: p=0.003) in ABCA4 carriers compared with the controls. Table 7 shows the means and SDs of the mfERG amplitudes in the right eye of the ABCA4 carriers and controls, and the p values of the Mann–Whitney U test results.
In autosomal recessive monogenic disorders, carriers of a single heterozygous mutation are expected not to experience any symptoms or signs of the disease, nor are they supposed to present with reduced function in any organ system. Despite this, heterozygous mutations in the ABCA4 gene have been suggested as a risk factor of AMD because the prevalence of certain ABCA4 mutations are overrepresented in patients with AMD, up to fivefold for some mutations [26,36]. Likewise, Fritsche et al. [37] described a certain subtype of AMD presenting with a fine granular pattern and peripheral punctate spots that are significantly associated with heterozygous ABCA4 mutations. Moreover, the prevalence of AMD has been shown to be significantly higher in the relatives of patients with STGD than in the general population, pointing toward a common genetic background [38,39]. Regarding retinal function, Maia-Lopes et al. [28] showed that the non-affected relatives of patients with STGD who were carrying a heterozygous ABCA4 mutation presented with reduced parameters concerning both electrophysiological measurements (mfERG) and tests using psychophysical approaches. Likewise, in a previous survey [27] conducted to study the effects of single specific ABCA4 mutations on retinal function and/or morphology, we investigated ABCA4 carriers and, surprisingly, found reduced macular function measured by mfERG. ABCA4 carriers were also found to have subtle and, in a few cases, more extensive morphological retinal changes on fundus photographs, fundus examinations, FAF and OCT. The aim of the present study was to conduct a 5-year follow-up of the same group of subjects to evaluate any changes of retinal status over time. As mentioned, the mfERG equipment, unfortunately, was updated during the 5-year follow-up. Therefore, no reliable comparison of the mfERG parameters could be made between the initial and follow-up measurement. When compared with the control group, the mfERG P1 amplitudes in the ABCA4 carriers were still reduced, and P1 implicit times were delayed in line with the results of the study of Maia-Lopes et al. [28]. Statistical analysis of the ffERG parameters, which reflect total retinal function and were recorded with exactly the same equipment on the follow-up and initial examinations, revealed a deterioration of the isolated rod function and combined rod and cone functions when responses at follow-up were compared with those at the initial assessment. Moreover, five subjects had reduced ffERG responses compared with the control group. Thus, over time, middle-aged ABCA4 carriers appeared to also experience a more general reduction of photoreceptor function, including rods, although they did not seem to notice it subjectively. A similar phenomenon with reduced general rod function has been demonstrated in patients with early AMD showing both delayed rod-mediated dark adaptation [40-43] and reduced rod function on ffERG [41,44]. In one study [42], visual acuity, low luminance acuity, macular light sensitivity, and spatial contrast sensitivity were normal, corresponding to the fact that all but two of our subjects did not notice any visual problems in daily life, although slight reduction in visual acuity was encountered in five of the 11 ABCA4 carriers. This reflects that electrophysiological measurement is more sensitive to incipient deterioration of retinal function than visual acuity testing. Forshaw et al. [45] also reported an association between macular dysfunction and reduced rod function in patients with AMD.
In the present study, 64% of the ABCA4 carriers showed morphological retinal alterations. The most prevalent findings were subtle pigmentary changes, quite similar to the pigmentary changes with hypopigmentations and hyperpigmentations observed in age-related maculopathy (ARM) and early dry AMD [46-48]. One subject also presented with drusen, which likewise can be associated with ARM and dry AMD [46-48]. When compared with those observed in the fundus photographs from the initial examination, retinal morphological changes at 5-year follow-up showed a subtle increase in 36% of the subjects. A similar slight increase in pathological alterations was also found in the FAF images. The most pronounced changes were found in Subject 8, who demonstrated widespread deep yellow flecks in the posterior pole, typical of STGD. The changes could now be explained by the fact that the subject was carrying the c.5603A>T variant in addition to the previously found c.768 G>T variant. c.5603A>T was not reported as pathogenic at the initial genetic workup and has been shown quite recently to be the missing ABCA4 variant in quite a few cases of late-onset ABCA4-associated retinal degeneration [34,35]. Subject 10, who might also be compound heterozygous for c.5603A>T and another ABCA4 variant (c.1804 C>T), (although we could not be entirely sure that they are in trans because no co-segregation analysis could be performed), had a less typical phenotype with only subtle pigmentary changes in the maculae and a few reduced electrophysiological parameters. Similar findings with weak penetrance and variable expression of ABCA4-associated retinal degeneration in patients with the c.5603A>T variant were reported by Runhart et al. [35]. Identification of missing alleles is of great importance in the management of patients with a phenotype that fits with ABCA4-associated retinal degeneration, but in whom previous genetic analysis failed to confirm the genotype. Our study further supports the significance of the c.5603A>T variant as disease causing, thus hopefully helping more patients to receive a reliable diagnosis, which is important for both genetic counseling and accessing future novel gene-based therapies.
A widely used explanation of the broad variation of the severity of the different forms of retinal degenerations associated with ABCA4 mutations (STGD, CRD, and arRP) is a model correlating the residual function of the ABCA4 protein to the extent of reduced retinal function [7,31,49-53]. In this model, on the basis of the calculated effect of the individual mutation on gene transcription and protein translation, null mutations, with no residual ABCA4 protein activity, are thought to be associated with the most severe forms of retinal dystrophies, and milder mutations only resulting in a defect protein are thought to cause milder entities. In line with this, carriership of a severe heterozygous ABCA4 mutation causing absent or extremely reduced residual ABCA4 protein transport function could lead to extremely poor clearance of toxic retinoid metabolites, resulting over time in reduced retinal function also in ABCA4 carriers, although carriers of autosomal recessive traits are not supposed to express the phenotype.
In the initial study [27], the c.768 G>T variant was considered a severe mutation associated with morphological retinal changes, marked macular dysfunction, and in one subject, also altered ffERG parameters. At follow-up, this pattern remained, still supporting previous studies classifying c.768 G>T as a rather serious mutation [3,54,55]. Concerning the c.2894 A>G, it seemed a bit milder in the initial study with normal ffERGs but still with moderate macular dysfunction in addition to OCT – and FAF alterations. The findings were similar at follow-up and in agreement with other studies describing c.2894 A>G as a quite serious mutation [54,56,57]. The remaining mutations were found in only one subject each; therefore, a thorough evaluation gets more uncertain. Nevertheless, c.319 C>T was still associated with a severe phenotype, and the subject with the c.4469 G>A variant still had a mild phenotype.
An important future perspective of a probable association between ABCA4 carriership and, at least, some AMD subgroups could be the search for novel treatment options for the dry form of the disorder, as there are ongoing studies on gene therapy for ABCA4-associated retinal degenerations [58-60] and pharmacological approaches [61]. One shortcoming of our study is that we failed to include a control group followed up for >5 years in the same way as the ABCA4 carriers. Thus, we cannot be entirely sure that aging is not a contributing factor to the reduced ffERG parameters over time.
Moreover, a larger number of subjects would have been advantageous, and finally, a comparison of mfERG parameters between the two time points would have been most interesting. Concerning the small number of subjects, the use of non-parametric statistical tests, to some extent, compensated for this. For the statistical analysis of mfERG results, we found it better to use a control group than to compare values obtained under non-equivalent conditions.
In conclusion, the 5-year follow-up of ABCA4 carriers, previously presenting with reduced macular function, revealed deterioration also of general retinal function with reduced rod, - as well as combined rod-cone amplitudes measured by ffERG in combination with slight increase of morphological changes in some subjects over time, which suggests that carriership of pathological ABCA4 mutations, may lead to a condition similar to a subgroup of AMD making this subgroup of AMD patients eligible for novel future gene-based – or pharmacological treatments. Moreover, we could confirm that in some cases of ABCA4-associated retinal degeneration with a missing allele, c.5603A>t testing could be useful for correct diagnosis; thereby, more patients can have access to optimized care.
We would like to thank Ing-Marie Holst and Boel Nilsson for skillful technical assistance. Funding: The work was supported by the Medical Faculty, Lund University, and grants from; The Armec Lindbergs Stiftelse; Stiftelsen för synskadade i f.d. Malmöhus län; Stiftelsen Synfrämjandets Forskningsfond/Ögonfonden. Availability of data and material: The authors have full control of all primary data and agree to allow the journal to review the data on request. Authors ‘contributions: Both authors made substantial contributions to the design of the work, the acquisition, analysis and interpretation of data as well as drafting and revising of the content of the manuscript. Both authors also approve the version to be published and take responsibility for all aspects of the work. Ethical approvals: The study was conducted in accordance with the Tenets of the Declaration of Helsinki and it was approved by the Ethical Committee for Medical Research at Lund University. Consent to participate: All subjects got a thorough written and oral explanation of the nature and possible consequences of the study and after that gave their written consent to participate. Consent for publication: All subjects gave their written consent to the publication of study results.