Article Text
Abstract
Background Birdshot chorioretinitis (BSCR) is a chronic bilateral posterior uveitis, which can affect central as well as peripheral vision. The aim of this study was to assess how visual acuity and visual field evolved over time in patients with BSCR.
Methods This was a prospective, observational, single-centre study based on data from the CO-BIRD cohort. Patient visits were categorised based on the time elapsed since the first symptoms, and groups of patients with different disease duration were defined. The main outcome measures were the best corrected visual acuity (BCVA), the mean deviation (MD) and the standard pattern deviation (PSD).
Results The study included 447 Caucasian patients (181 males and 266 females), all of whom HLA-A29 carriers. From onset to 30 years of disease duration, the number of patients in each consecutive 5-year period was 237, 250, 196, 147, 78 and 32, respectively. Overall, the range of visual acuity and visual field results increased with disease duration. BCVA gradually decreased and showed a significant decline after 11–15 years after the first symptoms. Among the visual field indices, PSD significantly increased after 16–20 years, while MD showed a significant decline after 21–25 years. No major gender differences were found in visual outcomes, indicating comparable severity. The intereye correlations of MD and PSD were stronger than those of BCVA.
Conclusions BSCR resulted in a large heterogeneity of visual outcomes, which increased with time. Our data provide an overview of the visual consequences of BSCR as a function of disease duration.
- Immunology
- Inflammation
Data availability statement
Data are available on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Birdshot chorioretinitis (BSCR) is a rare posterior uveitis, which can differentially affect the central and peripheral retina.
WHAT THIS STUDY ADDS
This study provides insights into the visual outcomes of BSCR and the coevolution of visual acuity and visual field testing over the long term.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Understanding the evolution and prognosis of BSCR is crucial to facilitate patient counselling, shape the focus of future research and treatment strategies.
Introduction
First described in the early 1980s, birdshot chorioretinitis (BSCR) is a chronic, bilateral posterior uveitis characterised by yellow-white lesions in the fundus.1 BSCR only affects the eye and has well-defined criteria for diagnosis and research.2 3 A study on BSCR in the UK reported an incidence of 0.035 cases per 100 000 person-years.4 BSCR is almost exclusively diagnosed in subjects of Northern European ancestry. From the immunogenetic standpoint, it is unique for its association with the HLA-A29 allele, which is the strongest link between an HLA class I antigen and any disease.5 As with other autoimmune diseases, the inflammatory activity associated with BSCR is long-lasting and the overall outcome varies according to the individual patients. Manifestations of BSCR include macular oedema, retinal vasculitis and papillitis of varying degree, which are sometimes accompanied by mild inflammation in the vitreous. These clinical signs can be different from one eye of a patient to the other, from one patient to another and in each patient at different time points during the disease. In addition, peripheral atrophy and other ocular complications can occur. Therefore, BSCR can affect the central and peripheral retina differently, resulting in a variable loss of central and/or peripheral vision. In this study, we analysed and compared the results of visual acuity (VA) and visual field (VF) testing over the course of the disease. Our goal was to have a global assessment of the vision of patients with BSCR over time.
Methods
This study was based on data from the CO-BIRD cohort (COhort of patients with BIRDshot chorioretinopathy—ClinicalTrials.gov Identifier: NCT05153057), which was conducted at a single institution (Hôpital Cochin, Paris, France). This prospective study, launched in 2002 and still ongoing, involves patients who are offered a yearly standardised evaluation. Our data collection methods have been previously reported.6–9 The disease diagnosis was based on the criteria defined by an International Consensus Conference and further confirmed by the Standardisation of Uveitis Nomenclature group.3
The disease duration was approximated by the time elapsed since the first symptoms. Six groups, each spanning a 5-year period, were used for the study and the patients’ VA and VF were assessed during 1–6 consecutive periods spanning up to 30 years. If a patient had multiple visits during a 5-year period, the data from the middle visit were used. When a patient had an even number of visits during a 5-year period, the data from the first of the two middle visits were selected. The main outcome measures were the best-corrected VA (BCVA), the mean deviation (MD) and the standard pattern deviation (PSD) over the course of the disease. The BCVA was measured with a decimal scale, converted to the logarithm of the minimum angle of resolution for the purpose of data analysis and was treated as a continuous variable. A BCVA of 0.01 and 0.001 (decimal notation) was assigned for a vision of ‘counting fingers’ and ‘hand motion’, respectively.10 The VF testing was performed using automated perimetry (Humphrey VF analyser—Zeiss-Humphrey, San Leandro, California, USA) with the Fastpac, full-threshold 30-2 programme. The two global indices were used to assess widespread sensitivity loss (MD) and localised sensitivity loss (PSD). Missing data were labelled as ‘unknown’ for each variable and displayed in the tables. In patients with advanced VF defects, for whom automated perimetry could not be performed, Goldmann perimetry was used and for the study purposes the MD was entered as −30 dB. This corresponded to the lowest value on the scale of sensitivity loss as measured by the Humphrey perimeter. Subsequently, a sensitivity analysis was performed to assess the potential introduction of a bias. To this end, every analysis was repeated after excluding the eyes for which MD was recorded as −30 dB (57 right eyes and 59 left eyes from 49 patients among several disease duration groups).
Descriptive data are presented as the mean±SD and median with IQR for continuous variables, and as counts with percentages for categorical variables. We analysed the median values of the BCVA, the MD and the PSD in the different groups. Furthermore, we assessed the correlation coefficient of BCVA, MD and PSD. Additionally, we investigated the intereye correlation of BCVA, MD and PSD of the right and the left eye on a per patient basis. We also recorded whether patients were phakic or had cataract surgery and analysed the impact of cataracts on the assessed parameters.
To investigate the percentage of patients with a preserved central vision and/or VF, patients’ eyes were classified in four categories: (1) decimal BCVA>0.75 (>20/25 Snellen equivalent) and MD >−6 dB; (2) BCVA>0.75 and MD <−6 dB; (3) BCVA<0.75 and MD >−6 dB and (4) BCVA<0.75 and MD <−6 dB.
Statistical analyses were performed by using statistical software R and RStudio V.9.1 (RStudio, Cary, North Carolina, USA). The study focused on the eye as the unit of analysis. Statistical significance was determined at a p value of 0.05. Because of skewed distributions in most continuous variables, non-parametric tests were used to assess differences between groups. Kruskal-Wallis and multiple comparisons by using Dunn’s test were used to compare medians of BCVA, MD and PSD in different groups. The Bonferroni correction was implemented to adjust the significance level and account for the multiple comparisons. A Pearson correlation coefficient was calculated for the relationship between continuous variables.
Results
The study included 447 Caucasian patients (181 males and 266 females) and 940 visits were analysed. Table 1 summarises the demographic features and the characteristics of all patients at the time of their last visit and grouped according to the years elapsed since the first symptoms. Females represented 60% of the cohort and the mean follow-up was 6.1±5.6 years. The mean age at the time of the first symptoms was 50±11 years and the mean age at diagnosis was 53±12 years. The time elapsed since the first symptoms related to BSCR ranged from 0 to 30 years (online supplemental figure 1). Depending on the duration of the follow-up, the patients contributed data to varying numbers of periods. Specifically, the number of patients contributing data to 1, 2, 3, 4, or 5 five-year periods were, respectively, 155, 181, 55, 32, and 23—none exceeding 5 periods. The patients with more than 25 years of disease were the ones with the longest time interval (6.7 years) between their first symptoms and the diagnosis of BSCR. Overall, the time between the initial symptoms and the diagnosis took longer in each of the six consecutive groups, extending from 0.9 to 6.7 years.
Supplemental material
The results of VA and VF testing over time are shown in figure 1. As the duration of the disease advanced, the heterogeneity of the results became more apparent, as indicated by increased SD. These SD values increased from 0.22 to 0.36 for BCVA, from 5 to 11 dB for MD and from 1.9 to 3.7 for PSD (online supplemental table 1). When compared with the 0–5 years group, the decline in BCVA was significant after 11–15 years after the onset of the first symptoms. The median MD remained stable at −3.7 dB for the first three groups spanning 15 years and a significant decline was observed after 21–25 years. The median PSD gradually increased with the disease duration and showed a statistically significant increase after 16–20 years after the initial symptoms. No significant gender differences were detected regarding these visual parameters except for higher MD and PSD in males in the 11–15 years group (online supplemental table 2). The Pearson R2 coefficient which assessed the correlation between BCVA and MD on a per eye basis increased from 0.14 to 0.49 over the course of the disease (figure 2). In contrast, the correlations between BCVA and PSD as well as MD and PSD exhibited more stability with the disease duration. The former generally showed lower R2 values around 0.11, while the latter showed higher R2 values around 0.49 (online supplemental figure 2).
The percentages of eyes with (>0.75) or without (<0.75) preserved BCVA and with (>−6 dB) or without (<−6 dB) preserved VF over the course of the disease are also shown in figure 2. Within the first 5 years after the first symptoms, 60% of the eyes had both preserved BCVA and preserved VF, but that percentage decreased over time to 24% after 26–30 years. Conversely, within the first 5 years, 14% of the eyes had both decreased BCVA and VF results, while that percentage increased to 46% after 26–30 years. Overall, VF testing showed more symmetrical results between the right and the left eye than VA measures, which was demonstrated by a Pearson R2 coefficient of 0.63 for MD and 0.37 for PSD, while the coefficient was 0.32 for BCVA (figure 3). These correlations are shown for all visits combined in figure 3 and detailed over the course of the disease in online supplemental figure 3.
A sensitivity analysis was performed to assess whether a bias was introduced by entering an MD at −30 dB in patients for whom automated perimetry could not be performed due to advanced VF loss. This analysis demonstrated that removing these eyes from the dataset did not result in major differences with our observed findings (data not shown).
Over the course of the disease, the percentage of cataracts ranged from 24% to 39% while the proportion of pseudophakic eyes increased from 12% (58/474) in the first group to 63% (40/64) in the last group (online supplemental table 3). The presence of a cataract primarily impacted BCVA and MD during the early stages of the disease (online supplemental table 4).
Discussion
The findings presented in this study are derived from the largest single-centre cohort of patients with BSCR. Our assessments encompassed patients at different time points over the course of their disease up to 30 years since their first symptoms. Early in the disease, VA and VF results were already heterogeneous, and the heterogeneity increased with the duration of the disease. VA loss progressed slowly over time and seemed to be gradual. On the other hand, median MD values remained stable for about 15 years after the onset of the clinical symptoms and then declined. Interestingly, PSD values showed a statistically significant difference earlier than MD values. This observation may be attributed to the fact that PSD is particularly sensitive to localised irregularities and subtle changes in the VF, which might manifest before the more widespread sensitivity loss captured by MD. The earlier detection of significant changes in PSD suggests its potential utility in identifying small VF alterations in the early stages of the disease.
Despite the higher frequency of women in our cohort, no major gender differences were identified regarding the assessed visual parameters over the course of the disease, suggesting a similar degree of severity.
In our study population, the prevalence of lens opacities remained fairly consistent across the disease duration groups, due to surgical intervention for significant cataracts. Furthermore, the stratification of data according to the presence or absence of a cataract revealed no impact on BCVA in later stages of the disease and vitritis was infrequent at these advanced stages. Additionally, the relatively strong correlation between MD and PSD suggests that often an increase in MD corresponds to an increase in PSD, indicating a localised loss independent of cataract. Taken together, these findings imply that the changes in BCVA and MD are likely associated with retinal damage. In the initial stages of clinically manifest BSCR, the disease is characterised by active inflammation, resulting in macular oedema as the principal cause of central vision loss. As inflammation subsides over time, vision loss can increasingly be attributed to the deterioration of photoreceptors’ inner/outer segments and/or epiretinal membranes. Decreased VF is presumed to be mainly linked to the progression of an atrophy of the mid-peripheral retina. Given that Humphrey VF testing does not encompass areas beyond the central 30°, our analyses might have missed further peripheral field defects. We only used Goldmann perimetry in patients with a decreased vision beyond the stage where automated VF testing could be performed. However, other studies of patients with BSCR have shown that Goldmann perimetry could be useful to assess the consequences of the disease on the further retinal periphery.11 In most eyes with vision loss, both VA and VF were concomitantly affected. Moreover, the correlation between VA and VF results increased as the disease progressed. Five years into the disease, 60% of the eyes maintained a good vision with both a decimal BCVA exceeding 0.75 and an MD above −6 dB. This proportion declined to 34% after a median of 22 years with a corresponding increase in the percentage of eyes showing declines in both BCVA and MD from 14% to 46%. However, we also observed eyes with a discrepancy between VA and VF results, either with a good central vision and significant field vision loss or the opposite.
Furthermore, for a given relatively good value of BCVA, there was a wide spread of MD values potentially due to localised defects within the VF. This notion is supported by a similar spread of the PSD for a given good BCVA (online supplemental figure 2).
To our knowledge, no prior reports have addressed the intereye symmetry of central vision and VF indices in BSCR. Our findings suggest a greater intereye symmetry for MD and PSD compared with BCVA.
We believe that the diagnosis of BSCR is made with less delay than in the first two decades that followed the description of the disease. Indeed, patients diagnosed more than two decades ago had the longest delay between their first symptoms and the diagnosis. Moreover, therapeutic progress, such as intravitreal corticosteroid injections, new immunosuppressants and biologicals, is now available,12 13 whereas these treatments could not be prescribed two decades ago. Hence, patients currently in the early stages of the disease are likely to have a more favourable outcome than patients diagnosed several decades ago.
In our study, certain biases cannot be neglected, such as the fact that patients with a longer follow-up were taken into account more often than patients with a shorter follow-up. We acknowledge that in some cases the recall of the onset of the first symptoms might be imprecise. However, our time interval groups spanned 5 years and we assume it is generally longer than the patients’ lack of certainty concerning the date of the first visual symptoms. In the case of advanced VF loss, Humphrey VF testing was no longer possible and Goldmann perimetry was performed instead (see online supplemental figure 4). To provide a global assessment of the VF, it was necessary to allocate these patients an MD of −30 dB. Long-standing inflammation, as well as regional and systemic treatment, can lead to a variety of ocular complications and secondary ocular diseases, such as cataract, epiretinal membranes, glaucoma, neovascularisation, etc. We acknowledge that these factors can have played a role in the overall analyses of our patients’ visual function. We did not investigate the effect of treatment since it was not the goal of this study. Treatment interventions can target macular oedema and/or papillitis and/or vasculitis assessed by fluorescein angiography. However, slowly progressing retinal atrophy of the mid-periphery might not be as easily detected as the inflammatory manifestations at the posterior pole, and therefore, may be more difficult to treat. Likewise, when it comes to monitoring treatment efficacy for atrophy, it is not as straightforward as assessing a reduction in central macular thickness measured by OCT in cases of macular oedema. This must be taken in consideration when comparing the evolution of VA and VF results.
Previous reports have highlighted the importance of monitoring both VA and VF in the follow-up of patients with BSCR.6 11 14–17 These studies were focused on visual function within a specific timeframe after presentation, after a defined follow-up period or disease duration. In a study involving 40 patients with a median disease duration at presentation of 3 (0–246) months, 10% of the patients had a better eye BCVA of 20/200 or worse.14 Patients with a disease duration of more than 30 months had higher frequencies of visual impairment to 20/50 or worse than patients who presented with a duration of disease less than 30 months (68% vs 32%).14 In another study of 98 BSCR patients with a follow-up period of at least 12 months, patients at presentation had a mean BCVA of 20/28.15 In a recent study involving 37 English BSCR patients, no difference was found when comparing BCVA between two groups of patients categorised based on the presence of first symptoms below or over a 32-week duration.16 Another study in the UK involving 46 BSCR patients who were monitored for a mean of 57.2 (5.8) months found stable BCVA and MD results throughout that follow-up period.17
Our study provides an analysis of visual outcomes based on a large single-centre cohort of BSCR patients, spanning three decades of disease duration. Our findings emphasise the gradual yet distinct decline in both central and peripheral vision as BSCR progresses. Our results underscore the wide spectrum and heterogeneity of BSCR’s visual outcomes. We also found that VF outcomes tended to be consistently more symmetrical between eyes than VA results. We will continue to investigate which factors influence the diverse visual outcomes of BSCR.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and the study was approved by the 'Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale' (ethics approval id 00011558) and maintained adherence to the Declaration of Helsinki for research involving human subjects. Informed consent was obtained from all participants before study enrolment. Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors JL and APB: study design, data analysis, interpretation. JT and DM: shaping key ideas and providing guidance on the analysis methodologies. DM, LI, SK and APB: conducted the ophthalmological follow-up of the patients and managed data acquisition. DM and APB: conceptualised, designed and supervised the study. APB is guarantor.
Funding This work was supported by the 'Association d’Ophtalmologie Cochin' (grant number: NA), the Bangerter-Rhyner Foundation (grant number: NA) and the Alfred-Vogt Foundation (grant number: NA).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.