Purpose
To identify potential predictors of permanent vision loss in patients with human leukocyte antigen (HLA)-B27-associated uveitis in a tertiary referral center.
Design
Retrospective case-control study.
Methods
The charts of 212 patients (338 eyes) with HLA-B27-associated uveitis that visited the University Medical Center Utrecht with a follow-up of at least 6 months were retrospectively studied. Clinical features at presentation and during follow-up were compared to final visual outcome in quiescent state. Eyes with (sub-) normal vision (>20/50) were compared with visually impaired (≤20/50) and blind (≤5/50, or a visual field of <10 degrees) eyes, using survival analysis. A multivariate Cox proportional hazards analysis was performed to analyze potential predictors for permanent vision loss.
Results
Median follow-up was 10.4 years (range, 0.5–44.7 years). During follow-up 226 eyes (66%) experienced vision loss up to 20/50, but most recovered. Twenty patients (9%) became permanently visually impaired or blind in at least 1 eye because of uveitis, after a median of 9.7 years (range, 0–20.9 years). The main cause was secondary glaucoma or related to glaucoma surgery (12/22 eyes, 55%). Survival analysis showed, after adjustment for age and sex, an ocular pressure of >21 mm Hg, hypotony, and panuveitis to be potential predictors at presentation, and the development of secondary glaucoma or hypotony to be predictors of blindness or visual impairment during follow-up.
Conclusions
The long-term visual prognosis of HLA-B27-associated uveitis is relatively good, but the true incidence of permanent vision loss is probably still underestimated. Our findings highlight the importance of proper control of intraocular pressure.
Acute anterior uveitis is the most common type of uveitis (intraocular inflammation) in referral centers in the Western world, and is associated with human leukocyte antigen (HLA)-B27 positivity in 6%–39% of cases. HLA-B27-associated uveitis typically comprises recurrent episodes of acute-onset, nongranulomatous anterior uveitis, usually limited to 1 eye or alternating between the eyes. Onset of uveitis is usually between the second and fourth decade, leading to high disease burden in a relatively young population. Patients are exposed to recurrent uveitis episodes, each time conveying the risk of complications and vision loss. Reported visual outcome is relatively good, with a surprising 0%–12.3% of all eyes developing a visual acuity of 20/50 or less. However, the risk of visual loss might be underestimated, since most of these studies have a follow-up time of less than 3 years.
Reported potential risk factors for visual impairment or blindness, including transient vision loss, include posterior synechiae at presentation, active inflammation, corticosteroid-sparing therapy, corticosteroid injections, chronic disease course, male sex, and posterior segment involvement. The aim of this study is to identify potential predictors of permanent visual impairment and blindness in a cohort of patients with HLA-B27-associated uveitis in a tertiary referral center.
Methods
Patient Selection
Medical charts of patients with HLA-B27-associated acute-onset uveitis that were seen at the Department of Ophthalmology, University Medical Center Utrecht (UMC-U) (Utrecht, The Netherlands), between January 1st, 1996 and December 31st, 2014 were retrospectively reviewed. All patients of 18 years and older with a minimum follow-up of 6 months and complete medical records were included. We excluded all patients in whom uveitis displayed highly atypical features, unlikely related to HLA-B27, such as an insidious onset of uveitis or posterior uveitis in the absence of inflammation of the anterior segment. This study was conducted in compliance with the Helsinki principles and federal laws in the Netherlands. Ethical approval was requested and obtained from the Medical Ethical Research Committee in Utrecht, with a waiver of informed consent acquisition because of the retrospective nature of this study.
Clinical Characteristics
Each patient underwent a full ophthalmologic examination by a trained ophthalmologist, routine laboratory screening, and a radiograph of the lungs. Laboratory screening included erythrocyte sedimentation rate, renal and liver function tests, angiotensin-converting enzyme (ACE), HLA-B27 positivity, and screening for various infectious agents in serum, including syphilis, Lyme disease, and an interferon-gamma release assay (IGRA) tuberculosis test. In case of atypical uveitis, aqueous humor was tested for infectious agents by means of polymerase chain reaction and Goldmann-Witmer coefficients.
For each patient the following characteristics were documented: age, sex, medical history including HLA-B27-related systemic diseases (ankylosing spondylitis [AS], reactive arthritis, psoriasis, and inflammatory bowel disease [IBD]), and date of onset of uveitis (defined as date of diagnosis by a [referring] ophthalmologist).
Uveitis was classified and graded in accordance with the Standardization of Uveitis Nomenclature (SUN) classification. For each eye the following features were recorded at presentation and during follow-up: cell grade in anterior chamber, presence of hypopyon, fibrin, posterior synechiae, corneal edema (defined as the presence of corneal thickening or Descemet membrane folds), keratic precipitates (KPs), intraocular pressure, duration of inflammation (limited: duration of <3 months; or persistent: duration of 3 months or more), course (acute: 1 episode with a duration of <3 months; recurrent: multiple episodes with a duration of <3 months; chronic: duration of ≥3 months and/or reactivity within 3 months after cessation of treatment; or quiescent under therapy: chronic treatment, not meeting criteria for chronic course). For each eye the following complications were recorded: posterior synechiae, band keratopathy, cystoid macular edema (CME), secondary cataract requiring surgery, ocular hypertension (defined as intraocular pressure >21 mm Hg without optic nerve damage or visual field abnormalities but requiring therapeutic intervention), secondary glaucoma (defined as ocular hypertension with optic nerve damage or glaucomatous abnormalities in visual field), and hypotony (defined as an intraocular pressure of <6 mm Hg on at least 2 measurements, and at 2 different time points, excluding postoperative hypotony within 3 months after glaucoma surgery).
Therapeutic interventions such as periocular corticosteroid injections, systemic steroids, disease-modifying antirheumatic drugs (DMARDs) (methotrexate, hydroxychloroquine, sulfasalazine, leflunomide, mycophenolate mofetil, azathioprine, cyclosporin, cyclophosphamide, tacrolimus, chlorambucil) and biologicals (adalimumab, etanercept, infliximab, rituximab, tocilizumab) were noted.
Vision
Best-corrected visual acuity (BCVA) was noted for each eye at presentation and when uveitis was quiescent, defined as no inflammatory cells in the anterior chamber during ophthalmologic examination. Also, the worst BCVA during follow-up was noted. Thresholds for visual impairment (VI) and blindness were used as recommended by the SUN working group, with a BCVA ≤20/50 (Snellen equivalent) for visual impairment and BCVA ≤20/200, or a visual field (VF) of ≤10 degrees, for blindness.
Visual acuities are presented in Snellen equivalent. For statistical analysis Snellen BCVA was converted into the logarithm of minimal angle of resolution scale (logMAR) and converted back to Snellen BCVA for data presentation. The most important cause of vision loss was noted for each eye, as well as the date that the threshold for blindness or visual impairment was first reached.
For statistical analysis eyes were divided into 2 groups: (sub-) normal vision (BCVA >20/50) or visually impaired/blind (VI/blind). Fourteen eyes could not be placed in either group: 5 eyes because their final BCVA was very likely to improve (because of persistent uveitis activity [n = 3] or preoperative cataract [n = 2]) and 9 eyes (from 8 patients) because their vision loss was predominantly caused by a non-uveitis-related ocular comorbidity. The respective causes of vision loss in these eyes are as follows: amblyopia (n = 2; the BCVA before uveitis was 20/40 and 1/300, respectively), Prader-Willi syndrome (n = 1), Fuchs endothelial dystrophy (n = 2), age-related macular degeneration (n = 2), and secondary corneal decompensation after surgery (n = 1) or after traumatic injury (n = 1). These eyes were included in the description of the group as a whole, but were excluded from comparison between VI/blind eyes and eyes that retained (sub-) normal vision, as well as from risk analysis. Eyes were also excluded from risk analysis if the criteria for permanent blindness/visual impairment were reached before presentation in the UMC-U, because these eyes were not “at risk” anymore, since the permanent blindness/visual impairment had already developed.
Statistical Analysis
Statistical analysis of the data was performed using SPSS version 21.0 (SPSS Inc, Chicago, Illinois, USA). To compare data between groups, Pearson χ 2 test, or Fisher exact test in case of <5 cases in 1 group, was used for categorical variables, and a Student t test or Mann-Whitney U test was used for continuous variables. A survival analysis was performed using Kaplan-Meier curves (with log-rank test) to correct for varying lengths of follow-up. To correct for paired samples only 1 eye per patient was used in the risk analysis: either the eye with the worst outcome (BCVA or visual field) was chosen, or the eye was selected randomly in case of similar outcomes. A univariate and multivariate Cox proportional hazards analysis was conducted. Variables were grouped as “present at presentation” or “developed during follow-up,” and multivariate analysis was conducted for these 2 time points separately. Variables with P < .05 in the univariate Cox proportional hazards analysis and potential confounders (sex, age of onset of uveitis) were entered into the multivariate analysis. P values below .05 were considered statistically significant. All significances were 2-tailed.
Results
The charts of 212 patients (338 eyes) with HLA-B27-associated uveitis, fulfilling the inclusion criteria, were analyzed ( Table 1 ). A total of 28 patients (13%) had significant permanent vision loss in at least 1 eye. This vision loss was directly caused by uveitis in 20 patients (9% of all patients) and was bilateral in 2 cases (1% of all patients). Eighteen out of the 22 eyes met the criteria for legal blindness and 4 eyes were visually impaired.
| All Patients | Permanent Vision Loss (Visual Impairment/Blindness) in ≥1 Eye | ||||||
|---|---|---|---|---|---|---|---|
| No a | Yes a | P Value | |||||
| N | % | N | % | N | % | ||
| Total number of patients | 212 | 179 | 20 | n.a. | |||
| Male sex | 112 | 53 | 96 | 54 | 10 | 50 | .76 |
| Follow-up years, median (range) | 10.4 (0.5–44.7) | 10.5 (0.5–44.7) | 12.2 (0.5–30.7) | .90 | |||
| Age of onset, mean (SD) b | 36.5 (14.5) | 35.9 (12.8) | 34.8 (17.5) | .81 | |||
| Age at referral, mean (SD) | 38.7 (13.6) | 37.9 (12.9) | 40.4 (15.1) | .41 | |||
| Associated systemic disease | |||||||
| Total | 104 | 49 | 91 | 51 | 8 | 40 | .36 |
| Multiple | 6 | 3 | 5 | 3 | 1 | 5 | – |
| AS | 87 | 41 | 77 | 43 | 7 | 35 | – |
| Reactive arthritis | 11 | 5 | 9 | 5 | 0 | 0 | – |
| Psoriasis | 8 | 4 | 6 | 3 | 2 | 10 | – |
| IBD | 4 | 2 | 4 | 2 | 0 | 0 | – |
| Therapy | |||||||
| Periocular steroids | 118 | 56 | 96 | 54 | 18 | 90 | .002 |
| Systemic steroids | 83 | 39 | 63 | 35 | 16 | 80 | <.001 |
| DMARDs | 47 | 22 | 36 | 20 | 9 | 45 | .02 |
| Biologicals | 31 | 15 | 21 | 12 | 8 | 40 | .001 |
| Surgery | 67 | 32 | 50 | 28 | 15 | 75 | <.001 |
a Fourteen eyes from 13 patients are not in either group: 3 eyes because of active inflammation at last follow-up, which resulted in a vision ≤20/50; 2 eyes because visual impairment resulted directly from (preoperative) cataract and postoperative vision is not known; and 9 eyes because vision loss is predominantly caused by an ocular comorbidity.
Ocular Signs at Presentation
The characteristics of all eyes at presentation and during follow-up are summarized in Table 2 . Thirty-two out of 337 eyes (9%) presented with panuveitis ( Supplementary Table , available at AJO.com ). In all cases of panuveitis the inflammation was unilateral and had an acute onset, and the inflammation of the anterior segment was more prominent than the posterior segment. In 12 of the 32 panuveitis cases (37.5%) the panuveitis was associated with an HLA-B27-associated systemic disease. In the other cases, extensive diagnostic examination was negative for other causes of uveitis.
| At Presentation | During Follow-up | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| (Sub-) Normal Vision a | VI/Blind a | P Value | (Sub-) Normal Vision a | VI/Blind a | P Value | |||||
| N | % | N | % | N | % | N | % | |||
| Total number of eyes | 302 | 22 | 302 | 22 | ||||||
| Episodes, median (range) | – | – | – | 3 (1–27) | 3.5 (1–20) | .87 | ||||
| Eyes with 3+ cells or more (%) | 42 b | 14 | 5 c | 24 | .21 | 144 d | 52 | 12 | 57 | .65 |
| Panuveitis | 26 | 8 | 2 | 18 | .13 | 58 | 19 | 9 | 41 | .02 |
| Laterality | ||||||||||
| Unilateral | 139 e | 48 | 13 | 59 | .04 | 53 | 18 | 8 | 36 | .03 |
| Alternating | 120 e | 41 | 4 | 18 | 176 | 58 | 7 | 32 | ||
| Bilateral | 33 e | 11 | 5 | 23 | 73 | 24 | 7 | 32 | ||
| Duration >3 months | 16 c | 5 | 1 | 5 | >.99 | 50 c | 17 | 10 | 46 | .002 |
| Course | ||||||||||
| Acute | 195 c | 65 | 15 | 68 | >.99 | 21 c | 7 | 3 | 14 | .09 |
| Recurrent | 87 c | 29 | 6 | 27 | 184 c | 61 | 9 | 41 | ||
| Chronic | 17 c | 6 | 1 | 5 | 87 c | 29 | 8 | 36 | ||
| Quiescent under therapy | n.a. | n.a. | n.a. | 9 | 3 | 2 | 9 | |||
| Fibrin | 50 | 17 | 4 | 18 | .77 | 121 | 40 | 11 | 50 | .36 |
| Keratic precipitates | 6 | 32 | 8 | 36 | .67 | 186 | 62 | 17 | 77 | .14 |
| Hypopyon | 6 | 2 | 1 | 5 | .39 | 17 | 6 | 5 | 23 | .01 |
| Corneal edema | 26 | 9 | 1 | 5 | >.99 | 52 | 17 | 12 | 55 | <.001 |
| Cystoid macular edema | 16 | 5 | 1 | 5 | >.99 | 48 | 16 | 13 | 59 | <.001 |
| Posterior synechiae | 75 | 25 | 7 | 32 | .47 | 171 | 57 | 19 | 86 | .01 |
| Hypotony f | 2 | 1 | 2 | 10 | .03 | 6 | 2 | 12 | 55 | <.001 |
| IOP >21 mm Hg f | ||||||||||
| Total | 15 | 6 | 9 | 43 | <.001 | 66 | 22 | 17 | 77 | <.001 |
| Ocular hypertension | – | – | – | 31 | 10 | 2 | 9 | >.99 | ||
| Glaucoma | – | – | – | 35 | 12 | 15 | 68 | <.001 | ||
| Cataract | 4 | 1 | 1 | 5 | .30 | 61 | 20 | 13 | 59 | <.001 |
| Band keratopathy | – | – | – | 2 | 1 | 3 | 14 | .003 | ||
a Fourteen eyes from 13 patients are not in either group: 3 eyes because of active inflammation at last follow-up, which resulted in a vision ≤20/50; 2 eyes because visual impairment resulted directly from (preoperative) cataract and postoperative vision is not known; and 9 eyes because vision loss is predominantly caused by an ocular comorbidity.
The VI/blind eyes presented significantly more often with an intraocular pressure of more than 21 mm Hg or less than 6 mm Hg than eyes that retained (sub-) normal vision (43% vs 6%, P < .001, and 10% vs 1%, P = .03, respectively). There was no difference between VI/blind eyes and eyes that retained (sub-) normal vision for severity of inflammation, defined as 3+ cells or more, or the presence of fibrin or hypopyon.
Ocular Signs and Complications During Follow-up
During follow-up, VI/blind eyes significantly more often progressed to panuveitis ( P = .02), the course was more often of a persistent duration ( P = .002), and more often a hypopyon or corneal edema developed ( P = .01 and P < .001, respectively). VI/blind eyes had developed more complications like CME ( P < .001), hypotony ( P < .001), secondary glaucoma ( P < .001), cataract ( P < .001), posterior synechiae ( P = .01), and band keratopathy ( P = .003).
Treatment
Patients with VI/blind eyes were treated more often with periocular injections, systemic steroids, DMARDs, and biologicals, and underwent more ocular surgeries ( Table 1 ). It is notable that patients with HLA-B27-associated systemic disease were significantly more often treated with DMARDs (34% vs 11%, P < .001) and biologicals (24% vs 6%, P < .001) than patients without a related systemic disease.
Visual Acuity
At presentation the median vision was 20/32, with 33% of all eyes having a BCVA of 20/50 or less. During follow-up 226 eyes (66% of all eyes) had a visual acuity of 20/50 or less at some point during follow-up ( Table 3 ). Fortunately, most of these eyes recovered.
| BCVA at Presentation | Worst BCVA | BCVA in Quiescent State b | ||||
|---|---|---|---|---|---|---|
| N | % | N | % | N | % | |
| Total | 335 | 338 | 281 | |||
| >20/50 | 224 | 67 | 114 | 34 | 252 | 90 |
| ≤20/50 to >5/50 | 64 | 19 | 137 | 41 | 11 | 4 |
| ≤5/50 to HM | 42 | 13 | 71 | 21 | 11 | 4 |
| LP + | 5 | 1 | 16 | 5 | 5 | 2 |
a Not including visual field abnormalities.
Visual Impairment and Legal Blindness
The primary cause of permanent vision loss is depicted in Table 4 . Secondary glaucoma or complications related to glaucoma surgery were the most frequent cause of blindness or visual impairment (13/20, 59%). Permanent vision loss developed after a median 9.7 years (range, 0–20.9 years) with a median of 4 (range, 1–20) episodes. Interestingly, the amount of uveitis episodes was not significantly different for VI/blind eyes compared with eyes with (sub-) normal vision.
| Cause | N | % |
|---|---|---|
| Secondary glaucoma | 12 a | 55 |
| Macular atrophy after longstanding CME | 3 | 14 |
| Serous retinal detachment | 2 | 9 |
| Fulminant inflammation b | 2 | 9 |
| Hypotony (ciliary body insufficiency) | 2 | 9 |
| Cornea decompensation | 1 | 5 |
| Total | 22 |
a In 2 out of 13 cases vision loss due to hypotony after glaucoma surgery.
b In 2 patients vision loss developed after fulminant inflammation with several coinciding complications. It is therefore impossible to assign 1 primary cause.
Eight eyes either were blind at referral (n = 2) or became blind after just 1 uveitis episode (n = 6). Five of these had had a severe panuveitis resulting in a combination of complications such as hypotony (n = 3), serous retinal detachment (n = 2), CME (n = 1), and optic disc edema (n = 1). Three other eyes that lost their sight after just 1 uveitis episode suffered from secondary glaucoma that aggressively progressed in the absence of inflammation.
Ocular Hypertension and Secondary Glaucoma
Eighty-nine eyes (26%) developed ocular hypertension, which progressed to secondary glaucoma in 51 eyes (57% of eyes with ocular hypertension, 15% of all eyes). The characteristics of glaucoma patients are summarized in Table 5 . Of these 51 eyes, 16 eyes (31%) developed severe visual field abnormalities such as a central scotoma (n = 1) or a double arcuate scotoma with >10 degrees of central field left (n = 7) or <10 degrees of central field left (n = 8).
| No Glaucoma | Glaucoma | P Value | |||
|---|---|---|---|---|---|
| N | % | N | % | ||
| Total number of patients | 173 | 39 | |||
| VI/blind in at least 1 eye | 13 | 8 | 15 | 39 | <.001 |
| Male sex | 95 | 55 | 17 | 44 | .20 |
| Follow-up years, median (range) | 10.1 (0.5–44.7) | 10.7 (1.2–44.3) | .78 | ||
| Age, mean (SD) | 49.8 (14.5) | 52.9 (15.9) | .24 | ||
| Age of onset, mean (SD) a | 36.9 (14.0) | 35.0 (16.6) | .50 | ||
| Age at presentation UMCU, mean (SD) | 38.2 (12.8) | 41.0 (16.7) | .27 | ||
| Follow-up outside of UMCU | 138 | 48 | 33 | 65 | .03 |
| Associated systemic disease | 87 | 50 | 17 | 44 | .45 |
| Therapy | |||||
| Periocular steroids | 85 | 49 | 33 | 85 | <.001 |
| Systemic steroids | 52 | 30 | 31 | 80 | <.001 |
| DMARDs | 28 | 16 | 19 | 49 | <.001 |
| Biologicals | 18 | 10 | 13 | 33 | <.001 |
| Surgery | 39 | 23 | 30 | 79 | <.001 |
| Total number of eyes | 287 | 51 | |||
| Episodes per eye, median (range) | 8 (1–27) | 5 (1–24) | <.001 | ||
| Duration >3 months b | 41 | 14 | 22 | 43 | <.001 |
| Chronic course b | 69 | 24 | 30 | 59 | <.001 |
| Laterality | |||||
| Unilateral | 54 | 19 | 13 | 26 | .001 |
| Unilateral alternating | 171 | 60 | 17 | 33 | |
| Bilateral | 62 | 22 | 21 | 41 | |
| Panuveitis | 47 | 16 | 23 | 45 | <.001 |
| Hypopyon | 17 | 6 | 5 | 10 | .30 |
| Fibrin | 110 | 38 | 26 | 51 | .09 |
| Cornea edema | 47 | 16 | 22 | 43 | <.001 |
| Hypotony | 10 | 4 | 9 | 18 | <.001 |
| Posterior synechiae | 160 | 56 | 41 | 80 | .001 |
| Cataract | 47 | 16 | 32 | 63 | <.001 |
| CME | 39 | 14 | 23 | 45 | <.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
