Introduction

Neuronal ceroid lipofuscinoses (NCL), collectively the most common progressive neurodegenerative disorders of childhood, comprise a heterogeneous group of lysosomal storage diseases [1–4]. These autosomal recessive syndromes result from mutations in at least 11 genes, encoding for either soluble lysosomal enzymes or transmembrane proteins (CLN1 through CLN10, and CLCN7) [4,5]. NCLs are grouped into congenital, infantile, late infantile (LINCL), juvenile (JNCL) and adult onset forms, depending on the specific gene mutations [1,4]. They are all characterized by an accumulation of hydrophobic autofluorescent material in the cytoplasm of neurons [4,6,7].

Mutations in the CLN2 gene, on chromosome 11p15, lead to classic LINCL, also referred to as Batten disease or Jansky–Bielschowsky disease [8–10]. CLN2 encodes the soluble CLN2 protein tripeptidyl peptidase 1 (TPP1), a lysosomal hydrolase that removes tripeptides from the N-terminus of small polypeptides [8,11,12]. CLN2 mutations result in an accumulation of curvilinear profiles in lysosomal residual bodies, the hallmark of LINCL [8]. The relationship between the accumulation of materials in the lysosomes and neuronal injury has yet to be fully elucidated [4,13].

Clinically, NCLs are characterized by developmental delay, psychomotor regression, ataxia, seizures, visual decline and premature death. The early clinical course of LINCL is dominated by seizures and ataxia, onset of which occurs around age 2 to 4. This is followed by cognitive and motor dysfunction, with complete loss of motor function by age 6 and death by mid-childhood [8,14,15]. In contrast to the early manifestations of vision loss in JNCL patients, visual impairment develops late in the disease course of LINCL, when progressive mental regression and gross motor disturbances have already become noticeable [4,8,16].

Vision loss in NCL occurs secondary to a progressive retinal degeneration of uncertain pathophysiology [4,8,16]. Previously reported ophthalmic manifestations, largely in JNCL patients, include a “bull’s eye” maculopathy, retinal pigment epithelial atrophy, fundus autofluorescence abnormalities, peripheral pigmentary disturbances, and a reduced or extinguished electroretinogram [17–25]. The ophthalmic manifestations of LINCL and their relationship with neurological function and the age of the patient have not been previously described.

In this current report, we characterized and categorized the ophthalmic manifestations of 25 LINCL patients with genetically confirmed CLN2 mutations. Based on these 50 eyes, we developed the Weill Cornell LINCL Ophthalmic Severity Scale and correlated the ophthalmic findings with the previously validated Weill Cornell LINCL Clinical Neurological Scale [15] and the age of the patient.

Methods

Results

Fifty eyes of 25 patients were included in the analysis (Table 1). All 25 patients had genetically confirmed CLN2 mutations and carried the diagnosis of LINCL. All 50 eyes underwent a comprehensive ocular exam along with FA and color fundus photography (Figure S1). Thirty-four of 50 eyes (68%) had an OCT, while 30 of the 50 eyes (60%) had an ICGA.

The mean age of the patients at the time of evaluation was 4.9 yr (median-4.8, mode -5.3, range 2.5 to 8.1). The mean LINCL neurological severity score was 6.1 (median -6.0, mode -5.0, range 2.0 to 11.0).

There were no anterior segment abnormalities noted in any of the 50 eyes examined. All 50 eyes had white and quiet conjunctiva, a clear cornea, deep and quiet anterior chamber, flat iris, and clear lens. The vitreous cavity in all eyes appeared normal with an optically clear vitreous. Posterior segment findings ranged from a completely unremarkable dilated exam, OCT, FA, and ICGA to a severe, widespread retinal atrophy with complete absence of the outer retinal structures and the retinal pigment epithelium (RPE, Figures 1-5). The retinal degeneration was first noted in the macula, specifically in the parafoveal region, with subtle outer retinal and RPE abnormalities (Figure 2). These abnormalities then progressed from subtle pigmentary changes in the fovea (Figure 3) to a fulminant bull’s eye maculopathy, with a more widespread involvement of the outer retinal structures in the macula (Figure 4). In our series of LINCL patients, there was 100% correlation between the ophthalmic score in one eye and the ophthalmic score in the contralateral eye. The Supporting Information (Figure S1) shows representative images obtained on all 25 patients.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Weill Cornell LINCL Ophthalmic Severity Score 1.

A. Dilated fundus photograph of patient 14 showing normal appearing optic nerve, vessels and fovea. B. Mid-phase FA and C. ICGA of patient 20 appear normal. D. SD-OCT of the same patient demonstrates normal retinal architecture without disruption of the outer retina. FA – fluorescein angiogram, ICGA - indocyanine green angiogram, SD-OCT – spectral domain optical coherence tomography.

https://doi.org/10.1371/journal.pone.0073128.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Weill Cornell LINCL Ophthalmic Severity Score 2.

A. Dilated fundus photograph of Patient 10 reveals subtle pigmentary changes in the fovea (arrow). The optic nerve and vessels appearing normal. B. Late-phase FA and C. ICGA of Patient 17 shows a faint area of central hyper-fluorescence (arrow) surrounded by hypo-fluorescence. D. SD-OCT of patient 10 demonstrates normal retinal architecture outside the fovea (arrows). E. Enlargement of the fovea and para-foveal regions of the same OCT exposes outer retinal abnormalities including the disruption of the ellipsoid hyper-reflective band (*). The external limiting membrane, however, appears intact. FA – fluorescein angiogram, ICGA - indocyanine green angiogram, SD-OCT – spectral domain optical coherence tomography.

https://doi.org/10.1371/journal.pone.0073128.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Weill Cornell LINCL Ophthalmic Severity Score 3.

A. Dilated fundus photograph of patient 7 reveals a central area of pigmentary change in the fovea (arrow). The optic nerve is somewhat pale in appearance with retinal vessels beginning to show signs of attenuation. B. Late-phase FA and C. ICGA of the same patient show a more prominent area of central hyper-fluorescence (arrow) surrounded by hypo-fluorescence. A few areas of pinpoint hyper-fluorescence outside this ring are also evident. D. SD-OCT of patient 7 demonstrates outer retinal abnormalities, including the disruption of the ellipsoid hyper-reflective band (*) as well as the external limiting membrane. This outer retinal atrophy is confined to the central fovea, extending less than 1 disc diameter. FA – fluorescein angiogram, ICGA - indocyanine green angiogram, SD-OCT – spectral domain optical coherence tomography.

https://doi.org/10.1371/journal.pone.0073128.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Weill Cornell LINCL Ophthalmic Severity Score 4.

A. Dilated fundus photograph of patient 9 shows a “bull’s eye” maculopathy with concentric rings of pigmentary changes emanating from the fovea (arrow). Some optic nerve pallor with attenuation of retinal vessels is also noted. B. Late phase FA of the right eye of patient 9 showing the “bull’s eye” maculopathy. C. Late phase FA of the left eye of patient 8 demonstrates a similar pattern within the macula. D. Late phase ICGA of the right eye of patient 9 correlates with the FA. E. Late phase ICGA of the left eye of patient 8. F. SD-OCT of patient 8 demonstrates considerable outer retinal abnormalities (*), including outer retinal atrophy and buildup of hyper-reflective material at the level of the retinal pigment epithelium. This outer retinal disruption typically extends less than 2 disc diameters (*), with normal appearing retinal architecture beyond the central fovea (arrows). G. Interestingly, the hyper-reflective material in some patients (including patient 16 shown here), appears to preferentially accumulate in the immediate parafoveal region (arrows). The late-phase FA (B and D) and ICGA (C and E) images show the hyper- and hypo-fluorescence areas (arrows) corresponding to the pigmentary changes on the color photograph (A) and the outer retinal abnormalities noted on the SD-OCT (F and G). FA – fluorescein angiogram, ICGA - indocyanine green angiogram, SD-OCT – spectral domain optical coherence tomography.

https://doi.org/10.1371/journal.pone.0073128.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Weill Cornell LINCL Ophthalmic Severity Score 5.

A. Dilated fundus photograph of patient 4 shows a “bull’s eye” maculopathy with concentric rings of pigmentary changes emanating from the fovea. The optic nerve is completely pale with severe attenuation of retinal vessels (arrow). B. In patient 19, with most severe neurological score in this series, the retina is completely atrophied with a pale optic nerve and barely recognizable retinal vessels. A “bull’s eye” maculopathy is not evident on the color photography in this severely affected patient. C. Late-phase FA and D. ICGA show a large “bull’s eye” maculopathy with concentric rings of hypo- and hyperfluorescence (arrow). Due to the retinal atrophy, the choroidal vessels are prominent throughout the ICGA (D, arrow). E. SD-OCT of patient 19 demonstrates considerable outer retinal atrophy with buildup of hyper-reflective material at the level of the retinal pigment epithelium. This outer retinal disruption extends throughout the entire fundus without normal appearing retinal architecture anywhere in the fovea. FA – fluorescein angiogram, ICGA - indocyanine green angiogram, SD-OCT – spectral domain optical coherence tomography.

https://doi.org/10.1371/journal.pone.0073128.g005

Correlation of the Ophthalmologic Score with Other Parameters

The mean LINCL ophthalmic severity score was 2.6 (median -2.0, mode -1.0, range 1 to 5, Table 1). An excellent correlation was found among the four retina specialists for determination of each patient’s LINCL ophthalmic severity score, where all correlations were highly significant (all Kendall’s tau > 0.87, p<0.0001). More severe ophthalmic manifestations were observed among patients with worse neurological findings (Kruskal-Wallis Test: X² = 4.8, p<0.03, Figure 6) and among older patients (Kruskal-Wallis Test: X² = 8.2, p<0.005, Figure 7). A direct correlation was found between this new Ophthalmic Severity Scale and the previously validated neurological scale (Kendall’s tau = -0.476, p<0.002, Figure 6). A direct association was also found between age and the Ophthalmic Severity Scale (Kendall’s tau =3.79, p<0.0002, Figure 7), with older patients having more severe ophthalmic manifestations.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Association of Weill Cornell LINCL Ophthalmic Severity Scale and Weill Cornell LINCL Neurological Score.

A direct correlation was found between this new Weill Cornell LINCL Ophthalmic Severity Scale and the previously validated neurological scale (Kendall’s tau = -0.476, p<0.002). More severe ophthalmic manifestations were observed in patients with more severe neurological findings (Kruskal-Wallis Test: X² = 4.8, p<0.03).

https://doi.org/10.1371/journal.pone.0073128.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 7. Correlation of Weill Cornell LINCL Ophthalmic Severity Scale and Age.

A direct association was found between age and the Weill Cornell LINCL Ophthalmic Severity Scale (Kendall’s tau =3.79, p<0.0002), with older patients having more severe ophthalmic manifestations (Kruskal-Wallis Test: X² = 8.2, p<0.005).

https://doi.org/10.1371/journal.pone.0073128.g007

Discussion

Neuronal ceroid lipofuscinoses are a heterogeneous group of lysosomal storage diseases characterized by a progressive neurodegenerative process resulting in loss of vision, seizures, movement disorders, mental regression, behavioral abnormalities, and a shortened life expectancy [1–4]. In the present study, we have systematically categorized the ophthalmic findings of LINCL and correlated the ocular abnormalities with clinical parameters.

We describe and classify the ophthalmic manifestations of LINCL in 50 eyes of 25 patients with genetically confirmed CLN2 mutations. The ocular findings are typified by a gradually progressive retinal degeneration, commencing at the level of the outer retina (specifically in the RPE and the photoreceptors, in a bull’s eye pattern) and progressing from the central macula to the peripheral retina. In the most severe cases, this retinal degeneration ultimately results in widespread retinal atrophy encompassing the entire fundus. Interestingly, none of the eyes in our cohort exhibited any anterior segment abnormalities, regardless of the severity of the retinal degeneration, the advancing age of the patient or the extent of neurological deterioration, which is in contrast to other lysosomal storage diseases that can manifest with anterior segment findings such as corneal opacities [26,27].

Based on the findings of the ophthalmic exam and the ancillary ophthalmic testing (including FA, ICGA, and OCT), we developed a five step Weill Cornell LINCL Ophthalmic Severity Scale (score of 1 to 5, from least to most severe retinal findings). Eyes with a LINCL score of 1 show no apparent ophthalmic abnormalities. This is in direct contrast to the eyes with a severity score of 5 that have extensive retinal degeneration and widespread retinal atrophy. The Weill Cornell LINCL Ophthalmic Severity Scale was easily comprehended and learned by our ophthalmology colleagues as evidenced by the high degree of correlation among the 4 independent masked graders.

The ophthalmic findings noted in our cohort of patients are consistent with the few previous reports of ocular manifestations of LINCL [17–25]. These descriptions of LINCL patients are typically part of larger NCL cohorts, and usually include the fundus examination with limited and variable ancillary testing. To the best of our knowledge, this current cohort is the largest series of LINCL patients who have had prospectively performed, standardized, comprehensive ophthalmic evaluations and neurological assessments, and the only cohort of NCL patients to have had ICGA and spectral domain OCT evaluations. Although the ICGA may add little in terms of diagnosis or classification of severity beyond what is noted on the FA, the progression of outer retinal atrophy and pigmentary changes is easily recognizable on the high-resolution spectral-domain OCT images. Even very early subtle changes can be demonstrated on the OCT, most impressively in the eyes with a LINCL ophthalmic score of 2 that show subtle parafoveal outer retinal changes. A dilated fundus exam and a macular high-resolution OCT may, in the end, be adequate to establish the extent and to follow the progression of retinal abnormalities in LINCL patients.

A previously published report on the electron microscopic findings in a patient with LINCL confirms our clinical findings that the retinal destruction in this disease starts at the photoreceptor and outer retinal levels and progresses from the macular area to the periphery [24]. The microscopic findings noted in this report are comparable to that seen on the high-resolution spectral-domain OCT images in our patients. The outer retinal (photoreceptor and RPE) atrophy is also similar to that noted in two reports of JNCL eyes evaluated at autopsy [17,19].

Our findings are also consistent with a prior description by Hainsworth and colleagues [25] of the funduscopic and angiographic appearance of NCL patients. In their entire cohort of NCL patients, Hainsworth noted optic nerve pallor, vascular attenuation, RPE atrophy, orange pigment clumping, and epiretinal membrane in the macula, along with bone spicules, RPE atrophy and course pigment stippling in the periphery. Of the 29 patients in the report, 24 were diagnosed with JNCL and 2 with LINCL [25]. These 2 LINCL patients were evaluated with color fundus photography alone. None of the 29 patients in this previous report had a systematic neurological assessment. As such, correlation of the extent of ophthalmic findings with NCL disease progression was not possible.

The optic nerve pallor, vascular attenuation, macular atrophy and pigmentary changes reported by Kelly and colleagues [22] in their cohort of 27 NCL patients are also consistent with our description of ocular findings in LINCL. In addition to color fundus photography, in this cohort, 9 NCL patients underwent standardized fundus autofluorescence (FAF) imaging. Kelly and colleagues [22] report an increased FAF early in the disease course and decreased FAF with more extensive retinal atrophy. The exact subset of LINCL patients who were imaged is not specified in the report, nor is the extent of disease progression noted in any of the NCL patients examined. The 25 patients in our current cohort did not undergo standardized FAF imaging, and as such, any possible FAF abnormalities were not incorporated into the Ophthalmic Severity Scale.

A distinctive aspect of our report is the integration of the extent of neurological worsening with the findings noted on ophthalmic evaluation. Not surprisingly, in the 25 patients evaluated in this study, there was a high degree of correlation between the Weill Cornell LINCL Ophthalmic Scale and the Weill Cornell LINCL neurological scale, with more progressive retinal degeneration noted in patients with worse neurological deterioration. All patients exhibited at least some extent of neurological dysfunction as determined by the Weill Cornell LINCL neurological scale. The youngest patient was the most neurologically intact, with the best neurological score and the best ophthalmic score. The oldest patient had the most progressive neurological deterioration in the entire cohort, and exhibited widespread retinal degeneration. Interestingly, Kendall’s tau for correlation between ophthalmic and neurological scores, while significant, was not exceedingly high, suggesting that the two scores may describe similar but not identical parameters. This may be due, in part, to the fact that the neurological score represents an aggregate of feeding, gait, motor and language abilities and thus defines a more global neurological deterioration noted in LINCL patients. The ophthalmic score, on the other hand, is specific to the progressive degeneration in a particular part of the CNS, that of the eye. Moreover, the ophthalmic testing (in particular the OCT, FA and ICG) may be more sensitive than gross motor and cognitive skills to subtle changes associated with the progressive degeneration in LINCL. It is also possible that the motor and language skills reflected in the neurological score have a slightly differing degenerative course compared with that of the ophthalmic findings reported here.

In a prior study, Steinfeld et al [14] developed a neurological rating scale to assess LINCL disease progression (e.g., neurological score). This scale had a visual ability, where a patient’s visual function (ranging from recognition and grabbing of an object to no light perception) was incorporated into the overall LINCL scoring system. Although this was a useful initial step, neither an ophthalmic examination nor any ancillary ophthalmic testing was a part of this study. To the best of our knowledge, there has been no previous attempt to correlate the progression of LINCL with that of the ocular findings that include a full ophthalmic examination and a battery of ophthalmic tests.

One limitation of our study is the lack of ophthalmic evaluations in awake LINCL patients. The already apparent mental regression and gross motor disturbances evident even in the least severely affected LINCL patients in our cohort precluded any accurate evaluation of visual function (e.g. visual acuity, visual field) or ophthalmic examination in an outpatient unsedated setting. We attempted to examine the first 4 LINCL patients in our outpatient ophthalmology clinic. However, we could not reliably perform visual acuity assessment or even a complete dilated fundus exam. Ancillary testing (OCT, FA, ICGA) in this setting was impossible in any of these 4 LINCL patients. Therefore, all 25 patients were examined and ancillary testing was performed under anesthesia. The visual acuity assessment in the Hamburg LINCL scale mirrors our experience with the difficultly in accurately determining visual function in these patients (the visual score in this scale ranges from recognition and grabbing at an object, uncoordinated grabbing, reaction to light, and finally no reaction to visual stimuli) [14].

Given the context of a gene therapy clinical trial from which all the patients were recruited, our cohort is biased toward less severely affected LINCL patients. As such, a majority of the eyes (28 of 50 eyes, 56%) in the current study demonstrated either no identifiable ophthalmic manifestations (Severity Score of 1) or very subtle changes that were noted primarily on ancillary ophthalmic testing (Severity Score of 2). These patients may be the ones who might ultimately gain the most clinical benefit from possible LINCL treatments for its ophthalmic manifestations, especially prior to the onset of severe widespread retinal degeneration. Moreover, response to any therapy for LINCL can easily be monitored in a relatively non-invasive fashion with serial ophthalmic evaluations.

As the patients who were included in this current manuscript were undergoing ophthalmic evaluations as part of a larger gene therapy trial, the availability of specific ancillary testing devices varied slightly between patients. The primary purpose of the ocular examinations was to screen for inclusion in the CNS gene therapy study rather than for the development of an ophthalmic scale. All patients were evaluated with a comprehensive dilated fundus exam, dilated fundus photography and FA. With the availability of specific equipment in the operating room, 68% of patients also had an OCT and 60% an ICGA. FA, ICGA and OCT each provide slightly different but complementary information about the retina. As typical for other ophthalmic disorders, abnormalities on OCT can also be observed on FA. The advantage of an OCT is that it is non-invasive, with no requirement for injection of a fluorescein dye to complete the testing. As LINCL changes appear to be predominantly retinal in nature, ICGA, typically utilized for detecting changes in the choroid, may not be important for the ophthalmic severity determination. Therefore, based on our experience, the most important ophthalmic examinations for evaluating the extent of disease severity included a dilated fundus exam and an OCT. FA and ICGA may not be necessary in all cases. There is, however, a possibility that the availability of OCT and ICGA for all the patients may have influenced the ophthalmic severity score reported in this current study.

In the current study, we describe the ophthalmic manifestations of LINCL, develop an objective and easily comprehensible ophthalmic LINCL severity scale, and correlate the ophthalmic manifestations with the extent of neurological decline and advancing age of patients with CLN2 mutations. This newly established Ophthalmic Scale may serve as an independent clinical indicator of LINCL severity and disease progression. Our multimodality approach of using a dilated exam, combined with FA, ICGA, and OCT and correlating these with the systemic manifestations may also be applied more broadly to characterize the ocular manifestations of other inherited lysosomal storage disorders with known ophthalmic findings (e.g., Niemann–Pick disease, Tay–Sachs disease, Sandhoff disease, and others). These ophthalmic manifestations may ultimately be beneficial in the assessment of novel therapeutic approaches for the treatment lysosomal storage disorders, such as stem cell and gene therapy, directed either systemically or locally at the central nervous system or the eye.

Supporting Information

Acknowledgments

We thank M Yeotsas, I Cao, C Hollmann, B Van de Graaf, Minhee Cho, and Jane Myung for technical assistance; and N Mohamed and DN McCarthy for help in preparing this manuscript.

Disclaimer: Original research / Access to data: The corresponding author had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis