A new classication for retinitis pigmentosa including multifocal electroretinography to evaluate the disease severity

Aim: To establish a useful and objective classication for retinitis pigmentosa (RP) to evaluate the disease severity. Methods: This is a retrospective cross-sectional study. Visual acuity (VA), visual eld width, ellipsoid zone width on optic cohorence tomography and multifocal electroretinography values were obtained from medical records of patients with RP. A scoring criterion was developed wherein each variable was assigned a score from 0 to 5 depending on its distribution. The cumulative score (from 0 to 20) was used to classify disease severity from grade 0 to 5. The scores were correlated with each other and the nal grade. Results: Data of 152 eyes of 92 patients who had the results of all tests were reviewed. The mean age was 41.2 years. The mean VA of the patients was 0.13±0.16 Snellen lines. The majority of patients had a VA less than 20/40 (88.6%), a visual eld smaller than 20° (78%), and an ellipsoid zone width smaller than 7° (84.4%). The majority of the patients (85.4%) were in advanced stage of the disease (Grade 4 and 5). Conclusions: We present a simple, objective and easy to use disease severity classication for RP which can be used to categorise patients and to evaluate and compare treatment results.


Introduction
Retinitis pigmentosa (RP) is a complex group of hereditary retinal disorders which causes degeneration of retinal photoreceptors. It has been reported that the disease has a worldwide prevalence of 1/4000 and it can be inherited with different patterns. The primary pathogenesis of RP is degeneration of photoreceptors, which leads to a narrowing of visual eld (VF), night blindness and deterioration of visual acuity (VA) ending with total blindness [1].
The natural history of the disease has long been studied using functional measures such as VA, VF and electrophysiological tests. Up to now, there is no de nitive curative treatment for patients with RP.
However in recent years new approaches including gene therapy and stem cell transplantation are being extensively investigated [2]. Signi cant advancements have been made in understanding the genetic pathogenesis of retinal diseases and the rst retinal gene therapy was recently approved by Food and Drug Administration (FDA) for the treatment of RP. Numerous other gene therapy trials are underway for various forms of inherited retinal disease [3]. The other treatment option, stem cell transplantation provides trophic support for neuroprotection and regeneration of damaged retinal cells through the secretion of neurotrophic factors iin retinal degenerative diseases and prevent progression. [4] To evaluate the therapeutic e cacy of these treatment options, it is important to measure the severity of the disease. As well as the clinical assessment and VA, the disease can be examined by a variety of tests like optical cohorence tomography (OCT), VF and electroretinography (ERG) [5].
In a recent study [6], the researchers designed a classi cation based on the ndings obtained from basic clinical examination and available diagnostic tests including VF and OCT. It is simple, easy to use and widely applicable to all patients. OCT is extensively used as a tool to monitor RP, since it can demonstrate outer retinal changes, particularly the changes in the ellipsoid zone (EZ), which are well correlated with functional measures [7][8][9]. VF most closely capture the patient's perception of visual impairment [10], however it is subjective and has high test-retest variability. Unfortunately, this classi cation does not include any electrophysiological tests which give more objective data in retinal diseases.
Electrophysiological tests such as full eld ERG (ffERG) and multifocal ERG (mfERG) provide an objective, functional measurement of the retinal function. They are useful in diagnosing RP and monitoring the long-term disease course for prognosis and treatment response. ffERG assesses nonselective global responses of the retina and it has not been able to reliably detect small progression, especially in the end stage of RP. Therefore we believe that ffERG will not be a proper test for a classi cation system. Multifocal ERG measures the innermost 30° of the retina and greatly improved the reproducibility of functional measurements compared to VF testing. VA correlated well with the amplitude of the central segment of the mfERGs, ring 5 amplitudes of the mfERG strongly correlated with the scotopic Ganzfeld ERG mixed cone-rod response amplitude and VF area. Also in advanced cases, reliable mfERG responses could still be recorded, even if the ffERG was not reproducible. Therefore mfERG would be a more appropriate test for an objective classi cation system. [11,12] The purpose of this study was to develop a phenotypic classi cation for RP which would provide a valuable and objective measure of disease severity for clinicians and researchers.

Study design and setting
This was a retrospective study conducted to establish a classi cation system to evaluate the severity of RP by using a phenotypic classi cation system. The study was performed in the ophthalmology department of a tertiary hospital in accordance with the Declaration of Helsinki, after obtaining the approval of the Ethics Committee of the University (2017/480, 13.10.2017). Written informed consent was obtained from all participants of the study.

Patients
After receiving a complete medical history including patient demographics, disease duration, inheritance pattern, the patients received a detailed ophthalmic examination including BCVA and intraocular pressure measurements, anterior segment evaluation with slit-lamp biomicroscopy, color fundus photography, OCT, VF and mfERG. BCVA was recorded with a Snellen chart at a distance of 3 meters. VF examination was performed by Humphrey VF analyzer device (Carl Zeiss Meditec AG Germany), program 30-2 was used for testing of each eye. Multifocal ERG was recorded on mfERG Vision monitor (Metrovision, France). The mfERG test was performed according to the International Society for Clinical Electrophysiology of Vision (ISCEV) guidelines [13]. During the mfERG evaluations, a matrix of 61 hexagons of the individual mfERG responses were generated, and these hexagons were grouped into ve concentric rings (15°) centered on the fovea. We recorded the average amplitude and implicit time of the rst positive wave (P1) in these ve rings. In this classi cation system we used amplitudes of positive P1 waves of 5 rings which show a strong correlation with the other parameters according to the previous study by Nagy et al. [11]. Responses were analyzed according to ring averages of P1 waves and responses over 5.0 nV were recorded as detectable [14].

Severity classi cation
Patients were evaluated regarding to the outcome measures of BCVA, OCT, VF and mf ERG tests. BCVA, VF width, ellipsoid zone (EZ) width on OCT, amplitudes of P1 wave of 5 rings were recorded for each patient and a scoring criterion was established for each variable varying from 0 to 5 depending on its distribution. The cumulative score (from 0 to 20) was used to classify disease severity from grade 0 to 5 (Table 1 and 2). We also evaluated the mean deviation (MD) value of VF and central macular thickness (CMT) measurements on OCT for each patient.
Severity classi cation was done according to the system designed by Iftikhar et al [9] and a modi cation was done for evaluating the results of mfERG. BCVA was given a score of 0 if vision was ≥20/20; 1 if it was <20/20 to ≥20/25; 2 if it was <20/25 to ≥20/32; 3 if it was <20/32 to ≥20/40; 4 if it was <20/40 to ≥20/100; and 5 if it was <20/100. VF diameter was given a score of 0 if it was ≥120° or better;

Results
Data of 152 eyes of 92 patients who had the results of all tests were reviewed. The mean age was 41.2 (between 20-69 ) years and 54.4 % of the study group were male. The mean age at onset of disease was 19.4 years and the mean disease duration was 22.3 (between 7-46) years. There was a family history in 56% of the patients. Among the patients with a family history autosomal dominant pattern was found in 28%, autosomal recessive pattern was found in 67% and X linked pattern was found in 5% of the patients.
Demographic data of the patients were shown in The majority of patients had a VA less than 20/40 (88.6%), a visual eld smaller than 20° (78%), and an ellipsoid zone width smaller than 7° (84.4%). Visual eld diameter and ellipsoid zone width both had a strong relationship with visual acuity ( Table 5).
The sum of all scores was used to determine the severity grade, as described in table 2. Figure 1 shows examples of different severity grades. Distribution of the grades were as follows: There were 2 eyes (1%) classi ed as grade 0, 4 eyes (2.6%) as grade 1, 5 eyes (3%) as grade 2, 12 eyes (8%) as grade 3, 23 eyes (15%) as grade 4 and 106 eyes (70.4%) as grade 5 ( Table 5). The majority of the patients (85.4%) were in advanced stage of the disease (Grade 4 and 5).
The score and the severity grade of the disease correlated strongly with VA, EZ, VF and mfERG score (p<0.05). The anatomical scores (EZ score and CMT) of the retina strongly correlated with the functional scores (VA, VF and mfERG score and MD) of the retina (p<0.05). . This led us the opinion that the EZ probably represents organised or densely packed photoreceptors and that there may be scattered or fragmented photoreceptors beyond the edges of the EZ that are alive and functioning [6]. Although this classi cation is simple and easy to perform, we believe that subjective measures like visual acuity and VF may incompletely demonstrate the patient's experiences of the daily life and disease severity.

Discussion
It is known that ERG is a gold standard test for evaluating RP because it is an objective and quantitative measure of global retinal function. Unfortunately, the test is di cult and time-consuming and can be extinguished in the early stage of the disease, when the central visual acuity is still entirely preserved [11]. Because the traditional ERG does not seem to be sensitive enough to indicate the condition of the central retina, other methods have been sought. The mfERG technique, which allows a highresolution mapping of the macular area of the retina seems to be a more promising method for detection of the remaining foveal cone function which can be detectable even in advanced stage of the disease. [12]. In another clinical study, researchers assessed central retinal function in patients with advanced RP using the mfERG. They reported that mfERG responses were recordable in at least one area in all successfully tested patients with advanced RP and nonrecordable ffERGs [18].
To the best of our knowledge this is the rst clinical study including mfERG in a classi cation system. We believe that including mfERG as a parameter would increase the value of the classi cation.
The study includes a large sample size and a wide range of patients in terms of age, sex, mode of inheritance and disease duration. We believe that this classi cation produces objective measure of disease severity and gives opportunity to compare the results of different treatment modalities.

Conclusions
In order to assess the severity of RP and compare the e cacy of new therapies, it is important to establish outcome measures that are both reliable and easy to evaluate. We believe that this classi cation is simple, produces objective data about disease severity and gives opportunity to compare the results of different treatment modalities.         Figure 1 Classi cation of disease severity from grades 0 to 5 using BCVA, VF, EZ on OCT and mfERG