Optical Coherence Tomography of the Macula in Congenital Achromatopsia

Abstract

purpose. It is known that symptoms of congenital achromatopsia are caused by the lack of functioning cones, but there are very few published data on histologic changes in the retina in these cases. This study was conducted to examine in vivo the anatomic structure of the retina of patients with achromatopsia.

methods. Fifteen eyes of eight patients with congenital achromatopsia and 18 eyes of nine control subjects were examined by optical coherence tomography. Radial 6-mm scans were taken of the macula. The thickness of the neuroretina was measured both automatically and manually. Measurements were taken at the foveola and at distances of 1.5 and 3 mm. Total macular volume and the retinal thickness in the nine ETDRS regions were also calculated.

results. In patients with achromatopsia, statistically significant reductions were found in total macular volume and in the thickness of the central retina. Remarkable differences were found between the results obtained from different methods of measuring retinal thickness. Automated methods underestimated retinal thickness compared with manual measurements.

conclusions. The structure of the macula in achromats differs from that in normal subjects. A possible reason for the structural alteration is the qualitative and/or quantitative disorder of the cone photoreceptors, as the morphologic change is most expressed in the foveola. The automated methods are not always suitable for measuring retinal thickness in the foveola. The structural changes seen in the central retina of the patients could provide useful information for future gene therapy.

Typical complete achromatopsia (i.e., complete achromatopsia with reduced visual acuity, mediated only by rods), also referred to as rod monochromacy (ACHM2: OMIM 216900; ACHM3: OMIM 262300; Online Mendelian Inheritance in Man; http://www.ncbi.nlm.nih.gov/Omim/ provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD), is a rare autosomal recessive inherited congenital retinal disorder. Its prevalence has been estimated at approximately 1:30,000 to 1:50,000. ^{1 2} The disease is characterized by photophobia, nystagmus, seriously reduced visual acuity, and a complete inability to discriminate colors. Electroretinography (ERG) recordings and psychophysical tests typically show a complete absence of cone function, supporting the original theory (the so-called pure rod theory) of Galezowski in 1868, who proposed that the visual functions in complete achromats are mediated wholly by the rods. ³  

Besides the numerous data from clinical examinations, there are few published data on the histologic changes in achromatopsia. These examinations are difficult to perform, due to rapid autolysis of the retinal cells. 

Harrison et al. ⁴ published histologic data of a 19-year-old patient with achromatopsia, who died through violence. The examination was performed only 40 hours after death, so extensive autolysis was seen all over the retina. However, a considerable thinning of the outer nuclear layer (ONL) was observed, with a marked reduction in the number of cone nuclei and imperfectly shaped squat, conelike units were seen in all parts of the retina. The photoreceptor layer (PRL) had almost entirely disappeared from the macular area. Some of these changes may have been due to autolysis, but the comparison of a normal eye fixed 72 hours after death showed a dense distribution of cones in similar portions of the retina. 

Falls et al. ⁵ reported the case of a 69-year-old patient with achromatopsia, whose right eye was enucleated because of acute congestive glaucoma, 4 months after the first glaucoma episode. In the histologic sections, atrophy of the retinal nerve fiber layer (RNFL) and the ganglion cell layer was seen. There was no evidence that the glaucomatous atrophy of the inner retina had involved the outer layers, where a marked decrease in the number of cones was observed. These cones were morphologically intact, except in the fovea, where cones with ectopic nuclei and irregular size and shape were found. ⁵  

Animal models of achromatopsia (CNG3 gene knockout mice) showed a rapid degeneration of the cone photoreceptors within the first year of life. ⁶  

Optical coherence tomography (OCT) is a modern, noninvasive, noncontact method for the in vivo examination of the retinal morphology. It was first used in the late 1990s, and became widely used. Based on low-coherence interferometry, OCT provides an axial resolution lower than 10 μm. ^{7 8}  

A recent publication from Barthelmes et al. ⁹ reports a highly altered foveolar structure but no change in the thickness of the central retina in patients with achromatopsia observed by OCT. 

Methods

There are 12 patients with previously diagnosed congenital achromatopsia registered at our department; however, only 8 of them participated in the study. 

The study adhered to the tenets of the Declaration of Helsinki. Informed consent was obtained from all the subjects after explanation of the nature and possible consequences of the study. The participants underwent standard ophthalmic examination. Best corrected visual acuity (BCVA) was measured by Snellen chart. Color vision was assessed by Ishihara plates and Farnsworth 15D Hue test. Full-field electroretinography (ERG) was performed according to ISCEV (International Society for Clinical Electrophysiology of Vision) standards ¹⁰ (RETIport System; Roland Consult GmbH, Wiesbaden, Germany). ERGs showed normal rod function, but in no case could any cone-specific response be detected. All these clinical findings were characteristic of congenital achromatopsia. 

The clinical diagnosis was confirmed by molecular genetic examinations in all cases; mutations were found in either the CNGA3 or CNGB3 gene (Table 1) . ¹¹  

Fifteen eyes of eight of our patients with congenital achromatopsia were examined at our department in 2004 and 2005 by OCT (Stratus OCT; Carl Zeiss Meditec Inc., Dublin, CA). One eye of a patient was not examined, because it was amblyopic and therefore absolutely unable to fixate. The results were compared with those of 18 eyes of nine healthy, age-matched control subjects. 

OCT examinations were performed with dilated pupils, by an experienced examiner. We performed the standard macular mapping using the built-in algorithm of the device. Six 6-mm-long radial scans were taken of the macula, centered on the foveola. The generally severe nystagmus and poor fixating ability make OCT difficult in achromats. Each scan was repeated until well-centered, good-quality images were captured from all the subjects. The foveolar centering was verified also by the infrared fundus camera. 

The overall structure of the macula was analyzed using different tools. Total macular volume (TMV) and the average retinal thickness (RT) in nine regions of the macula (defined by the ETDRS [Early Treatment Diabetic Retinopathy Study] ¹² ) were calculated by the built-in software (version 4.0.1; Carl Zeiss Meditec, Inc.). RT along a scan was also measured by the software (Retinal Thickness) by marking the inner (RNFL) and outer (RPE) border of the retina. 

Manual measurements were also performed along a scan by either analyzing the A-scans (a 6-mm scan consists of 512 A-scans) or using calipers. The A-scan analysis was performed only in images captured parallel to the retina. The peaks on the densitogram revealed the optically hyperdense layers of the retina: inner limiting membrane (ILM), external limiting membrane (ELM), and junction of the inner and outer segment of the photoreceptors (IS/OS) and retinal pigment epithelium (RPE), from the inside (vitreous body) to the outside (sclera). ^{13 14} We measured RT as the distances between the beginning (outer border) of the peaks presumably representing the ILM and the RPE (Fig. 1) . With calipers, the measuring crosses were placed on the upper border of the ILM and the RPE layers, and the distance between them was interpreted as RT (Fig. 2) . 

RT was measured in the foveola and at distances of 1.5 and 3 mm from the center by automated and two manual techniques. 

Statistical calculations were performed (Statistica 6.0; StatSoft Inc., Tulsa, OK) software. The distribution of the data was checked by Shapiro-Wilks W test. Parametric (two-sample t-probe) and nonparametric (Mann-Whitney) tests were used for the comparison. Statistical significance was set at P < 0.005. 

Results

A noticeable structural change was observed in the central retina of patients with achromatopsia. The alteration is most marked in the foveolar area; the widening of the photoreceptor outer segment layer (PR-OS) is missing there, as well as the hyperdense layer representing the OS/IS junction (Fig. 3) . 

The retina in the macula was found to be thinner and flatter in the patients than in the normal control subjects. The reduced thickness of the retina was most prominent in the parafoveolar region; the normally observable parafoveal rim was barely visible. 

Quantification of the changes showed that the retina of the patients with achromatopsia was significantly thinner than that of normal subjects in the ETDRS regions, except for the foveola. TMV was also significantly lower (Table 2) . 

Use of the automated RT analysis method obtained similar results. In the foveola, no significant difference was found between the control subjects and the patients, but at distances of 1.5 and 3 mm from the foveola, the RTs of the patients were significantly lower (Table 3) . 

Performing manual measurements, we measured slightly higher RT (compared with the automated methods) in all the observed regions, but mainly in the foveola. With both of these methods, the RT was lower in the patients than in the control subjects in all regions, including the foveola (Tables 4 5) . 

Discussion

OCT is a suitable method for examining retinal morphology in vivo, as has been reported in several studies. ^{7 8 15 16} In contrast to the functional and molecular genetic properties, the morphologic changes of congenital achromatopsia have been subject to much less research. There are only a few histologic reports published on the topic, ^{5 4} and the circumstances under which the examination was performed were not always optimal, thereby compromising the reliability of the results. 

Biel et al. ⁶ investigated the morphology of the retina of CNG3 knockout mice. These animals lack a cone-specific cation channel, similar to those of humans with achromatopsia. As a result of this missing protein, the phototransduction cascade is disconnected. The animals completely lack cone-mediated light responses, as demonstrated by ERG, whereas the rod responses remain. In these histologic examinations, the general structure of the retina was not found to have been affected significantly. The photoreceptor layer (PRL) thickness was normal and the rods morphologically intact. Regarding the cones, the CNG3-knockout mice showed progressive degeneration. In 2-month-old animals, there were only 10% to 20% of the normal numbers of cones, whereas in 8-month-old animals no cones were seen at all. 

Despite the progressive morphologic degeneration observed by Biel at al. ⁶ in animal models, human achromatopsia is regarded as a stationary disorder, because the clinical symptoms do not progress. 

Barthelmes et al. ⁹ reported results of OCT examinations performed in patients with achromatopsia and blue cone monochromacy. They measured no difference in the RT in the foveola, and when analyzing the structural changes, they found OCT a suitable method to distinguish achromatopsia from blue cone monochromacy. 

The purpose of our study was to examine in vivo the anatomic structure of the retina of humans with congenital achromatopsia. The retina of the patients with achromatopsia seemed flatter and thinner on OCT images that those of control subjects. The central part of the macula showed a marked structural alteration, as the parafoveal rim was missing and the foveolar morphology was also altered. The thinning of the retina in the parafoveolar regions implies the involvement of not only the cone photoreceptors themselves, but also the cone-related structures. 

Quantification of the changes showed a significantly reduced RT in the patients in the macular regions defined by the ETDRS, except for the foveola (Table 2) . TMV was also significantly lower in the patients compared with the control subjects. 

We also analyzed the images using the RT tool of the software. The thickness of the retina was measured in five positions (at the foveola and at ±1.5 and ±3.0 mm). In the peripheral parts of the scans, the patients’ retina was significantly thinner than that of the control subjects (Table 3) . In the foveola no significant difference was found between patients and control subjects; the retina of the achromats was even slightly thicker. This was surprising, as the patients’ retina seemed definitely flatter on OCT images. 

As reported previously, in some cases the built-in algorithm of the Stratus OCT software underestimates the RT. ¹⁵ The software measures the thickness of the retina between the inner borders of two hyperdense layers. The inner one is the retinal nerve fiber layer (RNFL), and the outer one is theoretically the retinal pigment epithelium-choriocapillary complex (RPE-CC). This outer hyperdense layer contains a thin layer inner to RPE, regarded as the junction of the inner and the IS/OS. This layer is most apparent in the foveola, as the thickening of the photoreceptor outer segments makes it more prominent. The measuring algorithm of the software does not distinguish this layer from the RPE-CC, leading to an underestimation of the RT. Ignoring the thickness of IS/OS and the OS produces a difference of approximately 30 μm in the foveola and approximately 10 μm elsewhere in the macula, in comparison to that in the retinal images of healthy control subjects. 

To avoid this inaccuracy of measurement, we also performed manual methods (A-scan-analysis, calipers), to analyze RT in patients and control subjects. The RT between the inner border of the inner hyperdense layer (ILM) and the inner border of RPE-CC was measured in this way (Figs. 2 3) . 

When the RT was measured by manual techniques, the retina in the foveola was found to be thicker in both control subjects (by ∼30 μm) and achromates (by ∼10 μm), compared with the measurements obtained with automated tools. 

The manually measured thicknesses were higher in all areas than those measured with the automated methods, particularly in the foveola. Compared with control subjects, the RT was lower in patients with achromatopsia, even in the foveola, whereas the automated methods showed no difference in the foveolar region. 

In contrast, Barthelmes et al. ⁹ found no difference between control subjects and patients with achromatopsia in the foveolar RT, even when analyzing the OCT images by the longitudinal reflectivity profile (LRP) method, which is similar to the A-scan analysis. Using the LRP method, they measured the RT from peak to peak, which may lead to the underestimation of the thickness due to the width of the peaks. The results of this group and those of ours can be explained by the different methods of the analysis. 

The OCT analysis of patients with achromatopsia shows a highly altered retinal morphology. These changes are most apparent in the foveolar and parafoveolar areas, presumably involving the cones and the related structures. Unfortunately, the resolution of Stratus OCT is not high enough to distinguish the histologic layers of the retina fully, or to analyze the changes at the cellular level. The new generation of ultra–high-resolution OCTs could help to solve these methodological difficulties. ¹⁴ Measuring the amount of macular pigments would also be an indicator of the integrity of foveolar cone photoreceptors. Unfortunately, none of the methods previously reported to be suitable for these measurements—heterochromatic flicker photometry, ¹⁷ multiwavelength scanning laser ophthalmoscopy, ¹⁸ autofluorescence attenuation, ¹⁹ and resonance Raman spectroscopy ²⁰ —are available at our clinic. 

Because the molecular genetic background of the disease has been identified in approximately 80% of the cases reported in the literature (Kohl S et al. IOVS 2001;42:ARVO Abstract 1745), ^{21 22 23} genetic therapy for patients with achromatopsia may be a viable option in the future. Successful gene therapy implies that structures of interest are anatomically present. ²⁴ It is therefore extremely important that the in vivo morphologic findings in the patients be taken into account before designing any sort of gene therapeutic intervention. 

 Submitted for publication September 29, 2006; revised November 27, 2006; accepted March 15, 2007.

 Disclosure: B. Varsányi, None; G.M. Somfai, None; B. Lesch, None; R. Vámos, None; Á. Farka s, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Clinical and Molecular Genetic Findings in the Patients with Achromatopsia Examined by OCT

Patients	Age (y)	Gender	BCVA	Color Vision	Photophobia	Nystagmus	Scotopic ERG	Photopic ERG/30 Hz	Mol. Gen.
1	4	Male	0.15/0.15	−	+++	+	+	−	CNGA3
2	9	Male	0.1/0.1	−	+++	+++	+	−	CNGB3
3	10	Male	0.1/0.1	−	++	+++	+	−	CNGB3
4	11	Female	0.1/0.1	−	+++	+++	+	−	CNGA3
5	12	Female	0.1/0.1	−	+++	+++	+	−	CNGB3
6	38	Female	0.1/0.15	−	++	+++	+	−	CNGA3
7	65	Female	0.2/0.2	−	++	+	+	−	CNGA3
8	67	Male	0.2/0.2	−	++	+	+	−	CNGB3

 Photophobia and nystagmus were graded as slight (+), moderate (++), or severe (+++). All patients lacked color discrimination (−). ERG+, preserved; ERG−, extinguished responses.

Figure 1.

 

Measurement of retinal thickness in the foveola by analyzing A-scans, in control subjects (left) and in patients with achromatopsia (right).

Figure 2.

 

Measurement of retinal thickness in the foveola with calipers, in control subjects (left) and in patients with achromatopsia (right).

Figure 3.

 

The (grayscale) OCT image of the central retina of a healthy control subject (top) and a patient with achromatopsia (bottom).

Table 2.

 

RT in the ETDRS Regions and the TMV

		Region	Control (n = 18)	Achromatopsia (n = 15)	P
RT (μm) in the ETDRS regions	1	Foveola	202.3 ± 13.8	201.8 ± 17.8	=0.927
	2	Inner upper	282.5 ± 10.8	220.5 ± 26.6	<0.001
	3	Inner temporal	263.5 ± 22.7	216.3 ± 24.0	<0.001
	4	Inner lower	281.5 ± 10.8	223.6 ± 25.6	<0.001
	5	Inner nasal	281.5 ± 13.2	228.1 ± 25.4	<0.001
	6	Outer upper	244.6 ± 16.3	206.1 ± 20.8	<0.001
	7	Outer temporal	227.1 ± 20.0	195.1 ± 19.1	<0.001
	8	Outer lower	231.2 ± 11.9	200.6 ± 20.4	<0.001
	9	Outer nasal	255.9 ± 15.9	223.1 ± 24.6	<0.001
TMV (mm3)		Macula	6.98 ± 0.36	5.93 ± 0.57	<0.001

 Measurements were made in normal control subjects and patients with achromatopsia, by the software of the device. Data are expressed as the mean ± SD; bold indicates statistically significant differences.

Table 3.

 

RT of Control Subjects and Patients with Achromatopsia, Using Automated Algorithm

RT (μm)	Control	Achromatopsia	P
Foveola	153.6 ± 14.1*	160.8 ± 23.2, ‡	=0.106
±1.5 mm	279.8 ± 19.8 , †	232.3 ± 28.1 , §	<0.001
±3.0 mm	226.2 ± 24.6 , †	202.6 ± 27.2 , §	<0.001

 Data are mean ± SD;

*   n = 54;

†   n = 108;

‡   n = 45;

§   n = 90. Statistically significant differences are in bold.

Table 4.

 

RT Measured Manually by Analyzing the A-scans

RT (μm)	Control	Achromatopsia	P
Foveola	186.7 ± 15.5 *	170.2 ± 21.1 , ‡	<0.001
±1.5 mm	296.1 ± 37.7 , †	245.8 ± 31.3 , §	<0.001
±3.0 mm	245.9 ± 22.3 , †	218.8 ± 28.6 , §	<0.001

 Data are mean ± SD;

*   n = 49;

†   n = 98;

‡   n = 42;

§   n = 84. Statistically significant differences are in bold.

Table 5.

 

RT Measured Manually with Calipers

RT (μm)	Control	Achromatopsia	P
Foveola	184.9 ± 13.2 *	171.5 ± 19.2 , ‡	<0.001
±1.5 mm	302.1 ± 21.6 , †	243.2 ± 27.9 , §	<0.001
±3.0 mm	242.9 ± 26.7 , †	217.0 ± 25.3 , §	<0.001

 Data are mean ± SD;

*   n = 54;

†   n = 108;

‡   n = 45;

§   n = 90. Statistically significant differences are in bold.

The authors thank all patients and family members for participating in the study and Erika Kovács and Ildikó Weiszer for helpful assistance.