Purpose
To determine the prognostic factors of long-term visual outcome in birdshot retinochoroidopathy (BRC).
Methods
Design : Retrospective case series. Study Population : Successive HLA-A29+ BRC patients whose latest visit was between May and August 2013 at a single tertiary center (Pitié-Salpétrière Hospital, Paris). Observation Procedure : Endpoint visual status (remission or deterioration) was determined for each patient based on clinical and ancillary data from the latest visit including optical coherence tomography (OCT), automated visual field (AVF), and angiograms. Main Outcome Measure : Epidemiologic, clinical, OCT, AVF, angiographic, and electrophysiological data at baseline were correlated to final visual status.
Results
Fifty-five patients were included. Mean observation period was 8 years (range: 0.6-23 years). Mean disease duration was 9.8 years (range: 1.2-32.7 years). Female-to-male sex ratio was 1.6:1. Factors of good visual prognosis (remission vs deterioration) included at baseline: late age of disease onset (49.5 vs 45 years, P = .05), presence of vitreous inflammatory reactions >2+ (35.9% vs 6.2%, P = .04), vascular leakage on fluorescein angiograms (FA) (44.4% vs 12.5%, P = .03), absence of macular pigment epithelium atrophy on FA (88.9% vs 62.5%, P = .05), and presence of macular edema on OCT (33.3% vs 6.2%, P = .04). Preserved electrooculography light peak and Arden ratio ( P = .06) and presence of choroidal spots on infracyanine green angiograms (80.0% vs 53.3%, P = .08) seemed associated with the best prognoses.
Conclusion
This study suggests a series of prognostic factors of long-term visual outcome in BRC. Keeping in mind the insidious evolution of the disease, knowledge of such prognostic factors should help tailor the treatment and monitoring of birdshot patients.
Birdshot retinochoroidopathy (BRC) is a rare form of posterior autoimmune uveitis that is mainly characterized by the presence of deep, hypopigmented ovoid spots in the fundus and nodular lesions at the choroidal level.
If the disease is indeed chronic, there have been a couple of reports supporting its relative self-limitation after resolution of the initial bout of inflammation, and this was mainly based on the stability of visual acuity (VA) over time. However, there is now evidence of the possibility of subclinical evolution toward retinal atrophy and blindness. Additionally, further analysis of a recent study suggested no evidence of strong correlations between VA and other descriptive variables of disease activity, among them electroretinograms (ERG) and visual fields.
Today a spectrum of treatment options for BRC, ranging from surveillance to highly aggressive corticosteroid and immunosuppressive treatments, are commonly employed, rather than targeted treatments based on the results of evidence-based medicine. Consequently, patient care and visual outcome appear to depend heavily on the attending physician’s opinion, which can vary depending on physician country of origin and degree of specialization. As a result, the need for prognostic factors to guide medical management becomes increasingly evident.
The objective of this work was to search for prognostic factors of long-term visual outcome to help adapt the treatment and monitoring of BRC patients in a more standardized fashion.
Methods
Institutional review board approvals for retrospective chart reviews were obtained commensurate with the respective institutional requirements before the beginning of the study. Described research was approved by the Ethics Committee of the French Society of Ophthalmology and adhered to the tenets of the Declaration of Helsinki. Fully informed consent was obtained for all patients. This hospital-based retrospective study reviewed the files of consecutive HLA-A29-positive BRC patients whose latest visit was between May and August 2013 at a tertiary referral center (Pitié Salpétrière Hospital, Paris).
For each patient, demographic data and medical and treatment history were recorded. Clinical parameters were collected at baseline for both eyes and included best-corrected visual acuity (BCVA) (decimal scale converted to logarithm of the minimum angle of resolution [logMAR]), quantification of anterior segment cells (laser cell flaremeter analyzer; Kowa FC 1000; Kowa, Tokyo, Japan), and vitreous inflammatory reaction ; and analysis of fundus features. Ancillary parameters included optical coherence tomography (OCT) central macular thickness (Cirrus OCT; Carl Zeiss Meditec, Inc, Jena, Germany), fluorescein angiograms (FA) and infracyanine green (ICG) angiograms, electrooculograms (EOG), and full-field (ff) electroretinograms (ERG), following the protocols of the International Society for Clinical Electrophysiology of Vision (WIN 8000F monitor; Metrovision, Perenchies, France) and automated visual field (AVF) parameters including mean deviation (MD) and pattern standard deviation (PSD) (24-2 and 10-2 programs; Zeiss-Humphrey, San Leandro, California, USA).
For each patient, endpoint visual status was defined and corresponded to 1 of the 2 following categories: remission or deterioration, based on clinical and ancillary tests from the latest visit ( Table 1 ).
| Category | Definition | N | % |
|---|---|---|---|
| Remission | Cure + complete remission + partial remission | 39 | 70.9 |
| Cure | Complete remission >5 years after treatment withdrawal | 1 | 1.8 |
| Complete remission | No direct signs of intraocular inflammation:
| 18 | 32.7 |
| Partial remission | Stabilized VA (<1 line loss) but persistent insidious activity in at least 1 eye:
| 20 | 36.3 |
| Deterioration | Clinical and/or subclinical deterioration in at least 1 eye that does not fit any other category | 16 | 29.1 |
a Vitreous inflammatory reaction was clinically determined based on the Standardization of Uveitis Nomenclature criteria.
Main Outcome Measure
Epidemiologic, clinical, OCT, AVF, angiographic, and electrophysiological data at baseline were collected and compared between the groups based on their final visual status.
Statistical Analysis
It was demonstrated previously that there is a strong correlation between both eyes for multiple parameters in BRC except for visual acuity (VA). As a result, data were analyzed by patient rather than in terms of eyes. In turn, quantitative variables were established based on the worst value between the right and left eye and for qualitative variables, presence of abnormal values in at least 1 eye was considered for each patient. Baseline quantitative parameters were described and compared between the 2 groups (remission vs deterioration) using the Mann-Whitney-Wilcoxon test, while comparisons for qualitative parameters used Fisher exact test. P values of less than .05 were considered statistically significant. Statistical analyses were performed using R 3.0.2 (R Development core team, 2013, Vienna, Austria).
Results
Baseline Characteristics
Fifty-five white patients were included. Female-to-male sex ratio was 1.6:1 (62.0% female, 38.0% male). Mean age at diagnosis was 49.1 years (range: 30.3-73.9 years). At baseline, mean disease duration was 2.8 years (range: 0-24.4 years). Mean follow-up duration was 8 years (range: 0.6-22.9 years). At endpoint, mean disease duration was 9.8 years (range: 1.2-32.3 years).
Medical history consisted of high blood pressure in 27.0% of cases, hypercholesterolemia in 16.4% of cases, type 2 diabetes in 5.4% of cases, autoimmune diseases in 5.0% of cases, and neurologic conditions in 7.3% of cases (1 case of meningioma sparing visual tracts, 1 case of idiopathic epilepsy, 1 case of noncompressive pituitary adenoma, and 1 case of congenital deafness).
Clinical and ancillary data at baseline are listed in Tables 2 and 3 . Birdshot lesions were initially seen in 72.0% of cases on fundus examination. Additionally, 58.0% of patients displayed optic nerve swelling while 81.0% showed vasculitis upon funduscopic examination. No significant cataract was noted initially. Overall, 25.0% of patients had macular edema on OCT (central macular thickness >250 μm or presence of “intraretinal cysts”). Regarding visual fields, foveal thresholds were in the normal range for all patients at baseline.
| Variables | Baseline | Endpoint | ||
|---|---|---|---|---|
| OD | OS | OD | OS | |
| Quantitative variables (mean ± SD) | ||||
| BCVA | 0.21 ± 0.25 | 0.29 ± 0.42 | 0.18 ± 0.3 | 0.32 ± 0.5 |
| Anterior chamber flare (ph/ms) a | 18.2 ± 31.3 | 23.7 ± 68.3 | 3.0 ± 2.1 | 4.0 ± 1.5 |
| Vitreous inflammatory reaction (+) b | 1.2 ± 0.6 | 1.2 ± 0.7 | 0.5 ± 0.3 | 0.5 ± 0.4 |
| OCT central macular thickness (μm) | 240.1 ± 114.6 | 251.7 ± 125.2 | 230.0 ± 54.0 | 236.0 ± 71.6 |
| 10-2 MD (dB) c | −4.7 ± 6.3 | −5.9 ± 4.1 | −8.4 ± 9 | −7.3 ± 6.9 |
| 24-2 MD (dB) c | −7.7 ± 7.3 | −7.9 ± 7.2 | −8.0 ± 8.0 | −6.8 ± 7.9 |
| Qualitative variables, n (%) | ||||
| Vasculitis on FA | 44/52 (84.6%) | 45/52 (86.5%) | 18/47 (38.3%) | 21/47 (44.6%) |
| Macular leakage on FA | 14/52 (26.9%) | 17/52 (32.7%) | 3/46 (6.5%) | 7/46 (15.2%) |
| Macular RPE atrophy on FA d | 9/52 (17.3%) | 8/52 (15.4%) | 12/47 (25.5%) | 12/47 (25.5%) |
| Birdshot spots on ICG e | 36/50 (72.0%) | 36/50 (72.0%) | 7/37 (18.9%) | 8/37 (21.6%) |
a Normal value <8 photon units/millisecond.
b Vitreous inflammatory reaction was clinically determined based on the Standardization of Uveitis Nomenclature criteria.
c Automated perimetry performed with the Humphrey visual field analyzer, using the 24-2 and 10-2 programs.
e On ICG: Birdshot spots are hypocyanescent during early and intermediate stages of the angiogram.
| Quantitative Variables (Mean ± SD) | OD | OS |
|---|---|---|
| Full-field ERG b-wave AMP (mV) a b | 67.8 ± 43.7 | 69.4 ± 43.3 |
| Full-field ERG b-wave IT (ms) a b | 40.7 ± 5.08 | 42.2 ± 4.7 |
| Full-field ERG b/a (AMP) ratio a b | 3 ± 1.25 | 2.9 ± 2.7 |
| 30 Hz Flicker b-wave AMP (mV) c | 70.6 ± 45 | 63.9± 42.3 |
| 30 Hz Flicker b-wave IT (ms) c | 29 ± 6.3 | 34.6 ± 13.6 |
| EOG Arden ratio a d | 159 ± 31.8 | 160 ± 36.6 |
| EOG light peak (mV) a | 713.5 ± 319.4 | 745.5 ± 346.2 |
a ERG and EOG followed the latest protocols recommended by the International Society for Clinical Electrophysiology of Vision.
b Photopic responses in scotopic environment were recorded for full-field ERG. Abnormal values for b-wave amplitude and implicit time were, respectively, <80 millivolts and >40 milliseconds.
c Abnormal values for 30 Hz Flicker b-wave amplitude and implicit time were, respectively, <100 millivolts and > 30 milliseconds.
Endpoint Characteristics
Characteristics at endpoint are displayed in Table 2 . In summary, visual acuity was stable overall during follow-up (no statistical difference between baseline and endpoint values) and no patent cataract was noted except in 3 patients who underwent cataract surgery. Prevalence of macular edema decreased at endpoint as compared with baseline (26.9% vs 6.5% OD and 32.7% vs 15.2% OS, respectively, for baseline vs endpoint). We also noticed that 5 patients, who presented with architectural modifications and decreased macular thickness at baseline, could recover normal retinal anatomy on OCT. Interestingly, alteration of mean deviation over time seemed more obvious on the 10-2 than the 24-2 AVF, regardless of the presence of macular edema. It was also observed that no neovascular membrane was diagnosed on OCT or FA.
During the follow-ups, 100.0% of patients received systemic corticosteroid treatments for a mean duration of 7 years (range: 0.3-16.6 years). Time interval between disease onset and initiation of systemic corticosteroids was on average 2.9 years (range: 0-24.2 years). Forty-two of 55 patients (76.3%) received immunosuppressive treatments (mycophenolate mofetil, azathioprine, cyclosporine, interferon alpha 2a,cyclophosphamide, and methotrexate in, respectively, 44.4%, 13.2%, 9.4%, 28.3%, 1.9%, and 1.9% of cases) for a mean duration of 4.8 years (range: 0.4-21.4 years), after an average of 5.2 years after diagnosis was made (range: 0.3-24.8 years).
Intermittent local steroid injections were administered punctually as adjuncts to systemic treatments to treat unilateral relapses. A total of 7.5% of patients received intravitreal injections while 18.8% received periocular injections of corticosteroids.
Endpoint Visual Status
Among the 55 studied patients, 39 (70.9%) remitted during the follow-up (1 cured, 18 in complete remission, 20 in partial remission), while 16 deteriorated ( Table 1 ).
Prognostic Value of Baseline Characteristics on Visual Evolution
Baseline characteristics according to final visual status are displayed in Table 4 . Follow-up duration was not found to be different between the groups (deterioration: median = 8.1 years [interquartile range (IQR): 5.8-11.9]; remission: median = 4.9 years [IQR: 3.1-11.1], P = .22]). Besides, patients in the remission group were slightly older at diagnosis than those in the deterioration group. In fact, early disease onset seemed associated with unfavorable visual outcomes, regardless of disease duration (deterioration: median = 45.2 years [IQR: 40.9-51] vs remission: median = 49.5 years [IQR: 45.3-54.6], P = .05). On the other hand, we found no significant difference in terms of past medical history or sex and final visual status.
