Reduced Electroretinogram Responses in Morphologically Normal Retina in Patients with Primary Hyperoxaluria Type 1

Abstract

Purpose

To describe ocular findings in individuals with primary hyperoxaluria type 1 (PH1), focusing on the correlations between retinal anatomy and retinal function. To characterize the retinal alterations that occur at different disease stages by evaluating individuals with diverse degrees of renal impairment associated with PH1.

Design

A cross-sectional study.

Participants

Patients diagnosed with PH1 based on clinical criteria and genetic testing, treated in the Pediatric Nephrology Unit of the Ruth Children’s Hospital, Rambam Health Care Campus, Haifa, Israel between 2013 and 2021.

Methods

The ophthalmological assessment included a slit-lamp biomicroscopy of the anterior and posterior segment or indirect ophthalmoscopy. Electroretinography was employed for assessment of the retinal function, and retinal imaging included spectral-domain OCT and fundus autofluorescence. A systematic evaluation of the disease stage was based on clinical criteria including physical examination, purposeful imaging (X-ray, echocardiography, and US abdomen), and laboratory tests as needed.

Main Outcome Measures

Anatomical and functional assessment of the retina in patients with PH1, and the relationship between retinal dysfunction and kidney impairment.

Results

A total of 16 eyes were examined in the study of 8 children ranging in age from 4 to 19 years. Four eyes (25%) showed normal structural and functional retinal findings, 8 eyes (50%) presented functional impairment in the absence of pathological structural findings, and 4 eyes (25%) had advanced retinal damage that manifested as significant morphological and functional impairment. There was no direct relationship between the severity of the renal disease and the severity of the retinal phenotype.

Conclusions

Subjects with PH1 present varying severity levels of the retinal phenotype, with possible discrepancy between the clinical retinal morphology and the retinal function noted on electroretinography. These findings raise questions about the molecular basis of the retinal manifestations in PH1. The presence of functional impairment in the absence of evident crystal deposition in the retina suggests that, in addition to oxalate crystal accumulation, other biomolecular processes may play a role in the development of retinopathy.

Primary hyperoxaluria type 1 (PH1) is a rare autosomal recessive disease caused by a deficiency of the liver-specific enzyme alanine-glyoxylate aminotransferase (AGXT), leading to increased plasma levels of oxalate.¹ Because oxalate cannot be metabolized and must be excreted in the urine, excessive levels of this metabolite prompt its binding to calcium, forming highly insoluble calcium-oxalate crystals that accumulate in the renal tubules. Patients with PH1 typically present in childhood or early adolescence with recurrent kidney stones and nephrocalcinosis, eventually progressing to further renal involvement with decline in the glomerular filtration rate.² Over time, increased serum and urinary oxalate levels lead to the deposition of calcium-oxalate aggregates in additional tissues and organs such as the bones, myocardium, and eyes.³ Nonetheless, the clinical manifestations in PH1 are heterogeneous and in severe cases, may include infantile oxalosis with end-stage renal disease (ESRD) manifested in the first weeks of life.⁴

In the eye, calcium oxalate can accumulate in the conjunctiva, episclera, sclera, ciliary body, and all layers of the posterior segment, according to histopathologic autopsy findings.⁵ The retinal manifestations of PH1 reported in the literature are limited and vary significantly between individuals,5, 6, 7, 8 ranging from mild perifoveal crystals with preserved visual function to macular crystals with severe retinal alterations, retinal pigment epithelium (RPE) hyperplasia, subretinal fibrosis, and chronic retinal edema leading to decreased visual acuity.^9,10 Severe ocular changes accompanied by progressive loss of retinal architecture were reported as possible outcomes in infantile oxalosis,^5,7,9 although no correlation between the genotype and the retinal phenotype was established.¹⁰

Here we describe the retinal manifestations of PH1 in patients with varying degrees of renal impairment. We used retinal imaging to define the retinal morphology and electroretinography (ERG) to assess retinal function. We propose a new biomolecular process that may play a role in the development of retinopathy in patients with excessive oxalate levels.

Methods

Study Population

In this cross-sectional study, patients were recruited from the tertiary Pediatric Nephrology department at the Ruth Children's Hospital of the Rambam Health Care Campus (Haifa, Israel). The research was carried out in accordance with institutional guidelines and the Declaration of Helsinki after approval by an institutional review board. Informed consent was obtained from the subjects (or their legal guardians) after explanation of the nature and possible consequences of the study. Included were subjects with PH1 diagnosed on the basis of clinical criteria and confirmatory genetic testing showing biallelic AGXT mutations manifesting all stages of renal disease severity. Patients with typical kidney disease but no molecular diagnosis were excluded from the study. A pediatric nephrologist performed a systemic evaluation, which included review of existing clinical records, a thorough medical history, clinical examinations, and proposed imaging such as a bone X-ray, echocardiogram, abdominal ultrasound, and laboratory tests to provide comprehensive clinical data on each patient. Chronic kidney disease (CKD) severity was calculated according to the Kidney Disease: Improving Global Outcomes guidelines.⁹

A detailed ophthalmologic examination included assessment of best-corrected visual acuity (when feasible), slit-lamp examination, ophthalmoscopy following pharmacologic pupils’ dilation, and retinal imaging including spectral-domain OCT and fundus autofluorescence (Spectralis HRA + OCT Laser Scanning Camera, Heidelberg Engineering). Based on the patients' ability to cooperate, whenever possible, electrophysiological testing was carried out in accordance with the protocols of the International Society for Clinical Electrophysiology of Vision.¹¹ Dark-adapted ERG after 30 minutes in the dark and flash-visual evoked potentials were performed using UTAS-Big shot (LKC Technologies ).

ERG

The ERG responses were recorded simultaneously from both eyes as previously described¹² according to the standards for clinical ERG set by the International Society for Clinical Electrophysiology of Vision.¹³ The system configuration consisted of a Ganzfeld light source having a maximal strength of 761 cd∗s/m² and a data acquisition unit (UTAS-3000 LKC Technologies). The ERG responses were recorded with corneal bipolar electrodes (Doran Instruments Inc) and an ear-clip as a ground electrode. Prior to ERG recordings, 1% cyclopentolate hydrochloride (Cyclogyl, Bausch & Lomb) and 2.5% phenylephrine hydrochloride (Ak-Dilate, Akorn) eye drops were instilled for full pupil dilation, and drops of benoxinate HCl 0.4% were administered for topical anesthesia. Light-adapted ERG responses were recorded first (background illumination of 30 cd/m²) using a single white flash (2.5 cd-seconds/m²). Following 30 minutes of dark adaptation, the scotopic ERG responses were recorded using full-field white light stimuli of different intensities with a maximum strength of 664 cd-seconds/m².

For ERG analysis, the amplitude of the a-wave was measured from the baseline to the first trough, and the amplitude of the b-wave was measured from the trough of the a-wave to the following peak. Corresponding measures obtained from 16 age-matched, systemically and visually healthy subjects were used to derive the range of normal ERG amplitudes, which served as control data.

Next, to quantitatively assess the change in the ERG, the relationship between the amplitude of the dark-adapted a-wave and the b-wave was determined for each response from each patient eye. The dependence of the b-wave on the a-wave was used as an index for the functional integrity of the retina. The b-to-a wave amplitude ratio remains unchanged by disorders that affect only the photoreceptors but is predicted to decrease when postphotoreceptoral elements such as the bipolar cells are affected.¹⁴ Nonetheless, in states of severely affected dark-adapted a-wave amplitudes, indicating significant loss of the rod photoreceptor function, a reduced b-to-a wave amplitude ratio can reflect predominately the response of the dark-adapted cone system.^15,16

Statistical Analysis

Locally estimated scatterplot smoothing was used to plot a smooth curve with a 95% confidence band based on ERG data points obtained from the control group, representing the expected b-wave value for a given a-wave. The same method was used to plot curves for study groups, facilitating the comparison of each patient to the control group. Correlation between 2 ordinal variables was assessed using Kendall’s tau coefficient.

Because of the small cohort, correlation between kidney disease and ocular manifestations could not be assessed with sufficient statistical power.

Results

The files of 16 affected individuals have been reviewed, 10 girls (62.5%) and 6 boys (37.5%) from 10 different families, including 2 pairs of siblings and 5 cousins. All patients were Arab-Muslims, and the parents were first-degree cousins in all cases. At the time of inclusion in the study, the patients' ages ranged from 6 months to 19 years, with an average age of 8 years (mean: 12, standard deviation: 4.97). After the initial screening examination, 2 patients were lost to follow-up, whereas 6 other patients were unable to complete the full ophthalmic examination and ERG testing due to unwillingness or technical difficulties and were thus excluded from further analysis. Overall, 8 patients were able to undergo full ophthalmologic assessment and were included in the study.

Systemic Manifestations

Kidney and systemic disease of varying severity was observed among our cohort of 8 patients (Table 1). At the time of enrollment, there were 2 patients who had normal kidney function, 2 patients who had been diagnosed with ESRD during early infancy, 1 who had been diagnosed with ESRD in adolescence, and 3 patients who had been diagnosed with CKD 1-3 between the ages of 1 month and 1 year. Two patients without ESRD had nephrolithiasis, and 1 had nephrocalcinosis (Fig S1). One patient underwent a liver transplant, and 2 patients underwent kidney and liver transplants. The demographics and severity of kidney disease are shown in Table 1. Other systemic manifestations included pathological bone fractures (Fig S2) and thyroid abnormalities.

Table 1.

Demographics and Staging of Kidney Disease

Patient No.	Gender	Age	CKD	Transplantation
1	M	4	ESRD∗
5	F	7	CKD-2
6	M	9	CKD-2	Liver (10 y)
7	M	11	CKD-3
8	F	17	CKD-0
9	F	17	CKD-0
13	F	12	ESRD∗	Liver and kidney (4 y)
15	F	19	ESRD	Liver and kidney (15 y)

CKD = chronic kidney disease; ESRD = end-stage renal disease.

^∗ESRD in early infancy.

Ocular Manifestations

Ocular findings in the 8 patients (16 eyes) for whom full assessment was attainable are summarized in Table 2. The retinal manifestations varied among subjects but remained largely symmetric in each patient. Based on the results of the ophthalmoscopic examination, OCT imaging and ERG assessment, the study subjects were divided into 3 groups: those with morphologically and functionally intact retinal findings, patients with normal morphology but impaired retinal function, and those in whom retinal findings were morphologically and functionally abnormal. Of note, the anterior segments of all patients were normal on clinical examination.

Table 2.

Ocular Findings

Patient No.	SD-OCT	FAF/SLO-IR	ERG
1	Abnormal	Abnormal	Abnormal
5	Normal	Normal	Abnormal
6	Normal	Normal	Abnormal
7	Normal	Normal	Normal
8	Normal	Normal	Normal
9	Normal	Normal	Abnormal
13	Abnormal	Abnormal	Abnormal
15	Normal	Normal	Abnormal

ERG = electroretinography; FAF = fundus autofluorescence; SD-OCT = spectral domain OCT; SLO-IR = scanning laser ophthalmoscopy infrared.

The ERG has been categorized as abnormal if the amplitudes were reduced and/or the configuration was abnormal.

Group 1: Normal Retinal Findings

Two patients (subjects 7 and 8; age 11 and 17 years, respectively) had normal findings on ophthalmological examination and retinal imaging (Fig 3). Specifically, no retinal deposits were visible clinically or on OCT and autofluorescence, and there were no pigmentary alterations. Consistently, the photopic and scotopic ERG responses showed normal amplitudes and configuration (Fig 3 and Table S3). When compared with healthy controls, the amplitude ratios between the b-wave and the a-wave were normal for both patients (Fig 4).

Figure 3.

Retinal imaging and representative scotopic electroretinography (ERG) response of patient 7 (left) and 8 (right) demonstrating normal retinal morphology and ERG responses. A, Near-infrared reflectance images of the right eye. B, Scaning laser ophthalmoscopy and (C) spectral domain OCT of the right eye. D, Dark-adapted ERG response for the same eye.

Figure 4.

The relationship between the amplitude ratio of the electroretinographic (ERG) b-wave and the a-wave. For each ERG response, the b-to-a wave ratio was determined. The data points from patient 7 (red triangles) and patient 8 (green squares) were plotted against measures obtained from 16 visually healthy control subjects (small gray dots). Locally estimated scatterplot smoothing was used to plot a smooth curve with a 95% confidence interval for the control (dashed grey line and grey shade, respectively) and for the study patients (dashed red line), demonstrating normal b-to-a ratio of patients in this group.

Group 2: Normal Morphology but Abnormal Retinal Function

In 4 patients (subjects: 5, 6, 9, and 15; age: 7, 9, 17, and 19 years, respectively; total of 8 eyes), the morphology of the posterior segment was clinically intact, with no crystalline deposits or other structural abnormalities noted in the retina, optic nerve, or retinal blood vessels (Fig 5 and Fig S6). In contrast, the ERG responses in these subjects were abnormal and presented an impaired configuration (Fig 5, Fig S6, and Table S4). The amplitudes of the photopic ERG responses were normal in all eyes except for slight reduction in one eye of patient 5, but the dark-adapted ERG was impaired with moderate attenuation of the a-waves and b-waves in patients 5, 6, and 15. Notably, the ERG configuration was abnormal in all eyes, with a mostly reduced amplitude ratio between the dark-adapted b-wave and the a-wave compared with healthy controls (Fig 7 and Table S4). We were unable to obtain visual acuities for most of the patients, but subject 15 had an acuity of 0.9 in both eyes.

Figure 5.

Retinal imaging and representative scotopic electroretinography (ERG) response of patient 5 (left) and 6 (right) demonstrating normal retinal morphology but abnormal ERG responses. A, B, Fundus autoflouresence of the right and left eye, respectively. C, D, Near-infrared reflectance images of the right and left eye, respectively. E, G, Scaning laser ophthalmoscopy and (F, H) spectral-domain OCT of the right and left eye, respectively. I, Scotopic ERG responses of 250 cd-seconds/m² (upper panel) and 664 cd-seconds/m² (lower panel) demonstrating in each subject electronegative waveform.

Figure 7.

The b-wave to a-wave ratio indicates decreased postreceptoral responses. For each electroretinography response, the b-to-a wave ratio was determined. The data points from each patient were plotted against measures obtained from 16 visually healthy control subjects (small gray dots). Locally estimated scatterplot smoothing was used to plot a smooth curve with a 95% confidence interval for the control (dashed grey line and grey shade, respectively) and for the study patients (dashed purple line). The dashed purple line of the study group is below the 95% confidence interval of the normal cohort, demonstrating that the responses from the patients in this group were characterized by a low b-to-a wave ratio.

Group 3: Abnormal Retinal Morphology and Function

Two patients (subjects 1, 13; age 4 and 12 years, respectively) were included in this group. Both demonstrated significant retinal morphologic changes noted as extensive subretinal fibrosis accompanied by diffuse crystalline deposits in each eye (Fig 8). Patient 1 had moderate decrease in the photopic ERG amplitudes and a moderate-severe decrease in the amplitudes of the scotopic a-wave and b-wave. Similarly, in patient 13, the photopic ERG responses were mildly reduced, but the scotopic responses were more severely affected, showing considerable reduction of the a-wave and b-wave amplitudes (Table S5). In comparison with visually healthy controls, the patients in this group exhibited mildly electronegative ERG configurations, i.e., selective reduction in the amplitude of the b-wave exceeding that of the a-wave, reflecting dysfunction of postphototransduction elements, as seen in conditions that affect mostly the inner retina but relatively spare photoreceptor function (Fig 9). Of note, 2 other patients who were excluded due to inability to cooperate on ERG testing, thus precluding measurement of the retinal function, demonstrated significant retinal morphologic changes with diffuse crystalline deposition similar to patients 1 and 13. Remarkably, they both had ESRD at early infancy.

Figure 8.

Retinal imaging and representative scotopic electroretinography (ERG) response of patient 1 (top) and 13 (bottom) demonstrating abnormal retinal morphology and abnormal ERG responses. A, C, Scaning laser ophthalmoscopy and (B, D) spectral-domain OCT of the right and left eye correspondingly, demonstrating significant retinal morphologic changes with subretinal fibrosis accompanied by crystalline deposits. E, Scotopic ERG responses demonstrating in each subject electronegative waveform. F, Color fundus photography demonstrating retinal calcium-oxalate deposits, retinal pigment epithelium hyperplasia, and subretinal fibrosis.

Figure 9.

The b-wave to a-wave ratio indicates decreased postreceptoral responses. For each electroretinography response, the b-to-a wave ratio was determined. The data points from patient 1 (orange triangles) and patient 13 (green squares) were plotted against measures obtained from 16 visually healthy control subjects (small gray dots). Locally estimated scatterplot smoothing was used to plot a smooth curve with a 95% confidence interval for the control (dashed gray line and gray shade, respectively) and for the study patients (dashed green line). The dashed green line of the study group is below the 95% confidence interval of the normal cohort, demonstrating that both patients in this group presented mostly responses with diminished amplitudes and subnormal b-to-a wave ratios. Of note, although b-to-a wave ratio on this graph may seem to reach normal values (the green dashed line crosses the dashed gray line), this is, in fact, due to low maximal responses that result in a left shift of the curve.

Correlation of Ocular Manifestations to Disease Severity

In this cohort of patients, no consistent correlation was observed between visual manifestations and the severity of renal illness. Among patients with clinically and functionally normal retina, 1 had normal kidney function (patient 8), whereas the other had CKD-3 (patient 7). Four patients with clinically normal but functionally impaired retinas showed diverse kidney function ranging from normal to ESRD. Nevertheless, all patients with infantile ESRD (N = 2) had significant retinal, morphological, and functional abnormalities, suggesting a tendency of these patients to develop more severe ocular pathology. This observation is strengthened by the fact that 2 patients with infantile ESRD who were not able to complete the ERG test, and were thus excluded from analysis, showed marked morphological abnormalities.

Discussion

Primary hyperoxaluria type 1 causes a wide range of retinal symptoms and manifestations, from normal morphology and function to severe retinopathy with widespread deposition of crystals and vision loss.^5,10 Previous histopathology and retinal imaging data revealed oxalate deposits in the posterior segment, with substantial interindividual variability, suggesting involvement of the inner retina, outer retina, and to a lesser degree, the RPE and sub-RPE regions.^9,17, 18, 19, 20 Because plasma oxalate levels rise only once the glomerular filtration falls below a certain threshold, but typically fall after supportive renal therapy and especially following final liver and kidney transplantation, the ophthalmologic manifestations might vary depending on the time points assessed.

In this study, we used ERG for objective assessment of the retinal function. Surprisingly, we found 4 individuals with no morphological abnormalities but subnormal ERG amplitudes. Although normal retinal function in the presence of calcium oxalate deposits has been recorded in the literature, this is the first indication of decreased retinal function in a clinically normal retina in PH1 patients. Small et al¹⁸ described 24 patients with PH, 8 patients with bilateral retinopathy, but ERG was not assessed, and the visual acuity provided the only functional parameter. They found that retinal manifestations were symmetrical and correlated with the systemic disease severity. Birtle et al⁹ described 68 patients with PH1, 12 of whom had infantile PH1 manifesting severe ocular alterations, whereas 56 patients with noninfantile PH1 manifested mild or no ocular alterations as assessed clinically and using multimodal imaging. Dulz et al²¹ described 2 siblings with PH1, one with infantile PH1 and severe retinopathy, and the other with less severe kidney disease and a normal retina, demonstrating the genotype–phenotype variability in PH1. Derveaux et al¹⁰ described 5 patients with PH1, 1 patient with infantile PH1 and severe retinopathy, 1 with normal retina but without an ERG test, 2 patients with no clinical signs of retinopathy and a normal pattern ERG test on first examination, while in follow-up, 1 of them developed mild crystalline deposits but the ERG remained normal. Interestingly, they described 1 patient with normal retinal morphology but reduced visual acuity, but no ERG has been performed.

As compared with healthy controls, the ERG configuration in group 2 was abnormal in all eyes, with a reduced ratio between the amplitudes of the dark-adapted b- and a-wave. A selective reduction in the b-wave is termed as an electronegative waveform, and in the presence of a normally sized a-wave, usually indicates inner retinal dysfunction occurring mostly in the rod system.²² In contrast, an electronegative ERG in the presence of decreased a-wave amplitudes can be explained by either co-existing injury at the photoreceptor and postphototransduction level, as well as by selective rod defect resulting in dominance of the cone “photopic hill.”^15,16,23 As a moderate attenuation of the dark-adapted a-waves was observed among the subjects in group 2, the reduced a-wave to b-wave amplitude ratio suggests widespread disruption with greater inner retinal impact.

The intriguing discrepancy between the impaired retinal function and the absence of retinal morphological abnormalities noted in our study suggested that additional molecular pathways may contribute to the pathogenesis of oxalate-related retinopathy even in the absence of crystal deposition. In a previous study by our group,²⁴ we have shown that oxalate can spontaneously form ordered fibrils with no associated calcium. The oxalate fibrils caused severe retinal cytotoxicity in cultured cells, and when injected intravitreally in rats, the fibrils induced retinal dysfunction noted as abnormal ERG responses presenting slightly electronegative patterns similar to those seen in patients with PH. It is thus suggested that retinotoxicity triggered by fibrillar oxalate assemblies could account for the retinal dysfunction seen clinically in PH1 individuals who do not have crystal deposition in the retina.

The treatment of PH1 has been improved with new drug approaches in recent years.²⁵ A well-tolerated and effective treatment option for reducing urine oxalate levels is RNA interference, such as Lumasiran (Oxlumo).²⁶ By reducing oxalate levels, crystal deposits may be less likely to develop in the retina, thus requiring other measurements in addition to morphological assessment for monitoring of the retinal status. In this cohort, 4 patients with normal retinal morphology had abnormal scotopic ERG amplitudes or abnormal ERG configurations. PH1 patients may benefit from ERG testing in the absence of crystals on clinical examination to detect early or subtle changes in retinal function. To assess the definite role of ERG in monitoring and in treatment decisions in HP1, more research and longitudinal ERG tests are needed.

There are a several limitations in this study. Only 8 of the 16 patients with PH1 were able to complete a thorough ocular examination, including the ERG. Even though 8 individuals are a significant number, considering the rarity of PH1, subgrouping patients based on retinal function and morphology and evaluating the correlation with the systemic disease status is thus limited. Furthermore, due to the lack of a follow-up period, this study is prone to time point measurement bias.

Conclusions

The morphological and functional retinal manifestations of patients with PH1 were studied. We discovered minimal intra-individual variability and substantial interindividual variability, a finding that is broadly consistent with the literature. We discovered that morphologically normal retinas may have reduced ERG amplitude and, more precisely, a particularly reduced b- to a-wave amplitude ratio, which may indicate mixed retinal dysfunction with inner retinal abnormalities. This discovery implies that crystalline deposits are not the only mechanism by which hyperoxalosis impairs retinal function. To further understand the mechanism of retinal impairment in hyperoxaluria, additional research is needed.

Manuscript no. XOPS-D-22-00232.