Elaborate Evaluation of Farnsworth Dichotomous D-15 Panel Test Can Help Differentiate between Best Vitelliform Macular Dystrophy and Autosomal Recessive Bestrophinopathy

Abstract

Introduction:

The colour vision in bestrophinopathies has not been assessed in detail so far. The aim of this study was to explore the extent to which distinct types of bestrophinopathies differ in regard to colour vision deficiencies using Farnsworth Dichotomous D-15 and Lanthony Desaturated D-15 panel tests.

Methods:

Both D-15 tests were performed in 52 eyes of 26 patients with Best vitelliform macular dystrophy (BVMD) and 10 eyes of 5 patients with autosomal recessive bestrophinopathy (ARB). Two methods were used for a quantitative assessment of the colour vision deficiencies: moment of inertia method and Bowman method. The following parameters were calculated: confusion angle, confusion index (C-index), selectivity index (S-index), total error score (TES), and colour confusion index (CCI).

Results:

The median value of confusion angle for all stages of BVMD fell into a narrow range around 62, indicating normal results. The median confusion angle value was 57 in ARB patients within a very wide range down to −82, indicating non-specific deficits. These differences were statistically significant. Significantly abnormal C-index and CCI values were found only in ARB patients, being 2.0 and 1.49, respectively. The majority of parameters of D-15 tests were independent of the visual acuity in both bestrophinopathies.

Conclusions:

Elaborate evaluation of the D-15 panel tests might help establish a differential diagnosis between different bestrophinopathies, as the pattern of the colour vision loss is different between BVMD and ARB. The quantitative parameters of colour vision tests in bestrophinopathies are independent of the visual acuity.

Introduction

Best vitelliform macular dystrophy (BVMD, VMD2, MIM153700) and autosomal recessive bestrophinopathy (ARB, MIM611809) are both caused by mutations in the BEST1 gene [1, 2] which encodes bestrophin 1, a protein involved in chloride channel function in the retinal pigment epithelium (RPE) [3, 4]. It has been postulated that an inhibited bestrophin function leads to abnormal fluid transport across the retina and secondary accumulation of coalesced subretinal deposits (vitelliform material) beneath the RPE of the central retina [5]. Clinically, bestrophinopathies are characterized by progressive visual loss of variable onset, as well as impaired central visual function – metamorphopsia and acquired anomalous colour discrimination [6, 7]. BVMD has a dominant mode of inheritance; it progresses through various clinical stages, leading to increasing central retinal dysfunction with age. The electrophysiological hallmark of BVMD is a reduced Arden index on electro-oculogram [8]. ARB, on the other hand, is caused by recessively inherited mutations in the BEST1 gene [9–11]. Here the phenotype is marked by autofluorescence changes, as well as an absence of the light rise on the electro-oculogram [10]. Colour vision testing has not been extensively evaluated in BVMD and ARB so far.

Clinical assessment of acquired colour vision deficiencies commonly involves arrangement tests that require the observer to arrange a set of samples that differ only in their chromaticity according to their perceived colour similarity [12, 13]. The Farnsworth-Munsell Hue 100 Test (FM-100) [14, 15] was simplified by using a reduced set of 15 samples with the Farnsworth D-15 Panel test [16] and its desaturated version (Lanthony Desaturated D-15 test) [17]. Both Panel D-15 tests differentiate among protan, deutan, and tritan defects by the angle of axes along which confusions are made. Arrangement tests are meant to assess the severity of the defect, are easy to administer, and can be used with naive subjects.

Vingrys and King-Smith [18] proposed an elaborated analysis of the FM-100 that can also be applied to the D-15. This analysis provides information on the direction and specificity of the errors and a measure of the total errors. Bowman [19], on the other hand, proposed a sum of the colour differences as a measure of the magnitude of the errors. We compare the two methods of quantitative assessment of colour vision in bestrophinopathies. The aim of this study was to investigate different subtypes of bestrophinopathies (BVMD and ARB) by means of colour vision differences and to illustrate the value of extended numerical evaluation of Panel D-15 tests.

Materials and Methods

Patients

The study was designed as a retrospective observational study. It was conducted at the Department for Hereditary Retinal Dystrophies at the University Eye Hospital of Tuebingen in Germany. Approval of the Ethics Committee of the Eberhard Karl’s University in Tuebingen has been obtained (126/2018BO2). Patient consent was not required as this study was based on publicly available data. The research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki.

A systematic review of the institutional database was performed to extract the data of the colour vision testing and Snellen visual acuity (VA) in patients with BVMD and ARB. The results of Farnsworth Dichotomous D-15 and Lanthony Desaturated D-15 panel tests taken at the same time were analysed.

Medical records of 62 eyes of 31 patients with BEST1 mutations confirmed with Sanger sequencing were analysed. There were 17 males and 14 females. None of the patients suffered from congenital colour deficiencies. There were 26 patients (52 eyes) with diagnosed BVMD and 5 patients (10 eyes) with features of ARB. VA was tested using Snellen charts and then converted into the logarithm of the minimum angle of resolution (logMAR) scale.

Fundus appearance, fundus autofluorescence, and optical coherence tomography results were used to distinguish between BVMD and ARB [10]. The stage of macular involvement of eyes diagnosed with BVMD was classified in accordance with Gass classification [20]: (1) previtelliform (normal macula or only slight RPE disturbance); (2) vitelliform (an “egg yolk” lesion); (3) pseudohypopyon (sinking of the yellow deposits inferiorly); (4) vitelliruptive (“scrambled egg” appearance with partial resorption of yellow deposits); and (5) atrophic.

ARB has been diagnosed based on the following features: hyper- and hypo-autofluorescent areas at the posterior pole and intraretinal fluid [9]. Electrooculography was performed in accordance with the guidelines of the International Society for Clinical Electrophysiology of Vision [21].

Assessment of the Colour Vision

Colour vision was assessed using the saturated Farnsworth Panel D-15 and the Desaturated Lanthony Panel D-15 test at a distance of 50 cm with near correction. Both tests were performed for each eye separately; the results of both eyes were taken into consideration, as there is often an asymmetry found between the eyes during the course of the disease [22].

The tests were performed by different examiners. The task is to arrange 15 coloured caps in their apparent natural colour order with reference to one fixed cap. The samples subtend 1.5° at 50 cm. The test is administered under Standard Illuminant C providing an illuminance of 250 lux. The results are transferred to a diagram on the scoring sheet. The test is failed if an observer makes two or more major crossovers that are parallel to an axis line on the scoring sheet. The axis of these lines determines the nature of the colour deficiency.

The online evaluation tool by Bela Török (https://www.torok.info/colourvision/) was used to analyse the colour vision results. A qualitative assessment of the D-15 test results was first performed. The results were classified qualitatively as either normal, nonspecific, tritan, deutan or protan.

Further analyses according to the Moment of Inertia method [18] and the Bowman method [19] were then carried out. The Moment of Inertia method calculated the following three parameters.

1. The confusion angle, a measure of the type of defect. Confusion angles between −25 and 25 are considered red-green defects; confusion angles between −70 and −100 and between 70 and 100 are considered blue-yellow defects. Defects with angles not falling within these zones are considered non-specific. Deutans generally have an angle of less than 0 and protans have an angle of more than 0.

2. The confusion index (C-index) is an indication of the severity of the colour vision defect. In perfect arrangement, C-index should be 1.0; C-index above 1.78 indicates a colour vision defect.

3. The selectivity index (S-index) provides a measure of how well errors parallel one another; in other words, it quantifies the specificity (polarity or randomness) of a cap arrangement. Total error score (TES) is a measure of the average colour difference between caps.

In the Bowman method, a single measure of error is known as the total colour difference score (TCDS), which is the sum of the colour differences between adjacent caps. Colour confusion index (CCI) was then calculated, which is the ratio of the actual TCDS to the normal TCDS. The value of CCI 1.0 is considered as normal and means a perfect arrangement of the test.

Statistical analysis was performed using STATISTICA 13.0 software (StatSoft, Krakow, Poland). All values were presented as means, medians, and standard deviations. The Mann-Whitney test was used to compare independent groups. Relationships were analysed using the R Spearman correlation coefficient. A value of p < 0.05 was considered statistically significant.

Results

Cohort

The mean age of the patients was 41 ± 12.67 years. Overall, there were 52 eyes diagnosed with BVMD and 10 eyes diagnosed with ARB. Eyes with BVMD were divided according to Gass [20] classification into 5 stages. There were 4 eyes classified as previtelliform stage I, 11 eyes with vitelliform stage II, 6 eyes with pseudohypopyon stage III, 14 eyes with vitelliruptive stage IV, and 17 eyes with atrophic stage V.

Qualitative Classification

A qualitative assessment of the colour vision in the BVMD group revealed 42 (81%) normal results (Fig. 1 on the right), 6 (11%) diffuse (non-specific, mixed), 2 (4%) with deuteranomaly, 1 (2%) result with tritanomaly, and 1(2%) with protanomaly. In 10 ARB eyes, there were 5 (50%) normal results, 2 (20%) diffuse, 2 (20%) tritanomaly (Fig. 1 on the right), and 1 (10%) protanomaly.

Fig. 1.

Examples of charts of a Dichotomous D-15 Panel performed by a patient with autosomal recessive bestrophinopathy (ARB) (on the left) and by a patient with Best vitelliform macular dystrophy (BVMD) (stage 3) (on the right). The major axes of confusion are marked in the patient with ARB it is the tritan axis, while in the patient with BVMD it is a close-to-normal result. The parameters obtained with the Moment of Inertia method (confusion angle, confusion index, selectivity index, total error score) and the Bowman method (colour confusion index) are presented on the left.

Quantitative Classification

The data were analysed quantitatively using two methods. First, for the saturated test, the Vingrys analysis showed a statistically significant difference (p = 0.04) between median values of confusion angles of saturated D-15 test between BVMD and ARB patients. The median angle value was found within a wide range of 56.8 ± 63.9 (Q1 = −12, Q3 = 62) in ARB patients; the median values for BVMD were very close to 62.0 ± 42.15 (Q1 = 62, Q3 = 62), indicating a normal result (Fig. 2). For desaturated D-15 test, there was no significant difference in the confusion angle between BVMD and ARB patients.

Fig. 2.

Median confusion angle values (vertical axis) for Best macular dystrophy values (BVMD) and autosomal recessive bestrophinopathy (ARB) (horizontal axis) for saturated (white) and desaturated (green) D-15 tests.

For the saturated D-15, significant differences (p = 0.02) were found between median C-index values of BVMD and ARB (p = 0.02). The median C-index was 1 in BVMD. The highest median value of C-index was found in ARB patients, which was 2 (Q1 = 1.06, Q3 = 2.27) (Fig. 3). There were no significant differences in the median values of S-index or TES between groups for D-15 test or the desaturated D-15.

Fig. 3.

Median C-index values (vertical axis) for BVMD and autosomal recessive bestrophinopathy (ARB) (horizontal axis) for saturated (white) and desaturated (green) D-15 tests.

There were also significant differences for the median CCI between BVMD (median 1.0; Q1 = 1, Q3 = 1.23) and ARB group (median 1.49; Q1 = 1.12, Q3 = 1.67) for the saturated D-15 test (p = 0.04) (Fig. 4). For the desaturated D-15 test, the differences were not significant.

Fig. 4.

Median values of colour confusion index (CCI) (vertical axis) for Best macular dystrophy (BVMD) and autosomal recessive bestrophinopathy (ARB) (horizontal axis) for saturated (white) and desaturated (green) D-15 tests.

Visual Acuity

The median logMAR VA was 0.22 in BVMD group and 0.64 in the ARB group (p = 0.01). The values of median VA were significantly different between groups (p = 0.0003) and were as follows: 0.00 ± 0.03 logMAR in previtelliform stage, 0.16 ± 0.09 logMAR in vitelliform stage, 0.33 ± 0.23 logMAR in pseudohypopyon stage, 0.36 ± 0.29 logMAR in vitelliruptive stage, 0.46 ± 0.6 logMAR in atrophic stage, 0.64 ± 0.25 logMAR in group with ARB. There was a correlation found only between CCI and logMAR VA (p < 0.04) (Table 1). Moreover, there were no statistically significant differences for the parameters of confusion angle, TES, CCI, S-and C-index for the Lanthony Desaturated test.

Table 1.

Correlation of all parameters of the D-15 test (CCI, colour confusion index; TES, total error score; CI, confusion index; SI, selectivity index) with logMAR visual acuity

Parameters	BVMD		ARB
	R	p value	R	p value
CCI	0.34*	0.04*	−0.48	0.28
Confusion angle	−0.16	0.37	0.18	0.69
TES	0.15	0.41	−0.50	0.26
CI	0.12	0.64	−0.20	0.80
SI	0.41	0.09	0.60	0.40

BVMD, Best vitelliform macular dystrophy; ARB, autosomal recessive bestrophinopathy; R, correlation coefficient; p, statistical significance.

^*Statistical significance.

Conclusion

Bestrophinopathies, including BVMD and ARB, usually manifest as a reduction in the VA in the first or second decade of life. However, these diseases show a large variation in expressivity [10, 22], also in colour vision.

In our study, BVMD eyes revealed a median value of confusion angle, which provides the principal axis of colour confusion, of 62 ± 42.15, indicating perfect cup arrangement within a narrow range. Values in ARB patients were lower and spread widely, indicating non-specific defects. Median values of C-index, which shows the degree of colour loss and quantifies the severity of the defect, were around 1.0 in all stages of BVMD and 2.0 in eyes of patients with ARB. A perfect arrangement yields a C-index of 1.0, whereas a C-index of above 1.78 indicates a likely colour vision defect. Moreover, CCI reflecting the number and type of transpositions made by the subject was abnormal only in ARB patients. Most of the parameters did not correlate with the VA.

In the present study, the majority of the patients (81%) with BVMD showed normal results of D-15 test, whereas only half of ARB patients showed normal results. Thus, performing sophisticated analysis of D-15 test results allowed us to find significant differences between BVMD and ARB. Both dystrophies are caused by mutations in BEST1 gene, yet they display different modes of inheritance and different phenotypes. The ARB phenotype is distinct from that seen in BVMD. It was identified in 2008 [9] and it is caused by compound heterozygous or homozygous BEST1 mutation. Patients are showing a diffuse irregularity of the reflex from the RPE, which are distinct from the extramacular vitelliform lesions described in Best disease. This is most clearly demonstrated by fundus autofluorescence imaging.

Colour vision depends on the normal functioning of cone photoreceptors and the normal processing of these signals in the retina. Bestrophinopathies are RPE-photoreceptor interface diseases. Photoreceptor/RPE complex represents one of the most physiologically active regions of the body. Colour vision data can provide information about the integrity of cone photoreceptors in the retina. Red-green channels in the retina use long- and middle-wavelength cone signals, while yellow-blue channels use short-wavelength cone signals, which are compared against the sum of long-and middle-wavelength signals [23]. It is known that in the central retina, short-wavelength cones are less numerous [24]. Colour assessment is particularly suitable for detecting changes caused by retinal disease, and there is a tendency towards greater yellow-blue loss [25]. There are studies reporting a correlation between chromatic sensitivity and disease progression in age-related macular degeneration [26].

In a study by Verriest and Caluvaerts published in 1978 [27], red-green deficits, classified as type I, were found in Best disease. Similar findings were reported in Stargardt dystrophy, Leber hereditary optic neuropathy, optic neuritis, and toxic amblyopia [15]. No single mechanism encompasses all these disease processes. Acquired defects are often hard to classify and may be a combination of types. Individuals with acquired defects often have an associated decrease in VA and visual field loss. Unlike congenital defects, they appear after birth and may fluctuate in severity and type.

There are a large number of studies devoted to assessing colour vision deficits in congenital colour vision defects. However, only few studies address colour vision deficits in hereditary retinal degenerations. Boon and colleagues [11] found that patients with ARB were deficient in the blue-yellow axis along with one individual experiencing red-green confusion. In a recent study by Pfister and colleagues [28], 6 patients with ARB tested for colour vision failed, and three ARB cases (50%) tested using the Panel-D-15 test demonstrated tritan axis deficits. Pfister and colleagues postulated that the null effect of ARB mutations negatively impacts rods more than that seen in BVMD, resulting in a clinically significant effect that might be used to distinguish ARB from BVMD.

In a study describing phenotype-genotype correlations in an Australian cohort of BVMD patients, colour vision testing was undertaken using Ishihara’s test. Colour vision results were available for 30 (72%) mutation carriers and 12 (85%) clinical VMD patients. A total of 12 mutation carriers (28%) were unable to perform the test because of severe visual loss, being too young, or for some other reason. Total 15 (35%) mutation carriers were in group 1 (all Ishihara colour plates correct in either eye), 9 (21%) were in group 2 (between 20 and 24 plates correct), and 6 (14%) were in group 3 (less than 20 correct). Five clinical VMD patients were in group 1 (35%), five in group 2 (35%), and two in group 3 (14%).

In cone dystrophies, loss of colour discrimination parallels VA loss. In a clinic-based study performed by Thiadens [29], three commonly used colour vision tests: Hardy-Rand-Rittler (HRR); Lanthony, Farnsworth Panel D-15; and Ishihara tests were compared among patients with cone dystrophy, controls with normal VA, and controls with low vision and legal blindness. All tests have a high discriminative accuracy to identify cone disorders, but the highest scores were for the HRR test. Poor VA slightly decreased the accuracy of all tests [29].

Colour vision has also been investigated in Stargardt disease − the most common hereditary macular dystrophy [30]. Different tests were used: pseudoisochromatic Ishihara and Hardy-Rand-Rittler (HRR) plates, D-15 tests, and Nagel anomaloscopy. Normal colour vision was found in only approximately 20% of patients. Twelve of the 26 patients who failed the test only showed an R/G defect on the D-15 (46.2%). Contrary to the HRR test, 7 of the 26 subjects presented with an isolated B/Y defect (26.9%). Five subjects had both R/G and B/Y defects (19.2%). A scotopization pattern was seen in 1 patient (3.8%). The degree and type of colour vision deficiency in Stargardt disease patients correlated better with BCVA than with full-field electroretinogram results.

Colour vision has been assessed in a large cohort of patients with Usher syndrome [31]. A significant difference in the pattern of colour vision loss in USH1 and USH2 patients was found but not in the severity of the loss.

The Bowman method of scoring yields an index of severity in the cup arrangement. It has been already applied in acquired losses of colour vison as age-related macular degeneration [32].

TCDS and CCI increase with the number of trans-positional crossovers which indicate poor colour discrimination. However, in our study, the differences in these parameters were not statistically significant.

Low S-index values indicate that there is a high degree of randomness in the errors, whereas high values indicate that the errors are highly regular around a common axis of confusion. In subjects with congenital deficits, C- and S- indices were highly correlated [33].

It should be mentioned that both eyes of each patient were analysed separately in this study, as we have not taken into account intraclass correlation coefficient [34]. If the correlation is close to one, then data from both eyes could be averaged or one eye selected at random for analysis using conventional statistics. It can be considered as a limitation of this study.

Clinical colour vision testing takes the place of longer and more complicated psychophysical measurements [35], as there are factors that affect test performance like lighting, field size, as well as cognitive and emotional issues. In addition, most colour vision tests are designed for individuals with a certain level of cognitive ability and education [15].

In conclusion, colour vision tests should be performed as additional examinations in the clinical assessment of patients with macular dystrophies, including BVMD and ARB. However, it should be emphasized that the colour vision test performance should not be overestimated, and it is very unlikely that detailed colour vision testing would be used for diagnostic differentiation in routine clinical practice.

Data Availability Statement

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.