Abstract
Background/objectives
The predictive factors of surgical results in uveitic retinal detachment (RD) are lacking. The objective was to study the surgical outcomes and determine the risk factors for surgical failure in rhegmatogenous RD associated with intraocular inflammation (RRDIOI).
Methods
Retrospective series of consecutive eyes with RRDIOI undergoing vitreoretinal surgery between 2012 and 2019 in two French referral centres. Patients underwent 23- or 25 G pars plana vitrectomy (PPV), scleral buckling (SB), or a combination of both. The main objective was to describe the predictive factors of visual recovery and anatomical success after surgery.
Results
Seventy-one eyes were included. Posterior and panuveitis accounted for 91.5% of eyes. Seventy-five percent of eyes had an infectious cause for their uveitis. The first surgery consisted in PPV alone, SB alone, or both in 87.3%, 4.2% and 8.5% of cases respectively. The reattachment rate was 74.6% after one surgery (100% in case of SB, either alone or in association with PPV). On multivariate analysis, the only predictive factor of visual improvement was a baseline BCVA ≥ 20/400, while the only predictive factor for surgical success at 12 months was the absence of RD recurrence within the first 6 weeks of surgery.
Conclusions
RRDIOI has a relatively favourable anatomical prognosis. The addition of scleral buckling may be beneficial in selected cases.
Subject terms: Retinal diseases, Prognostic markers
Introduction
Retinal detachment (RD) is a sight threatening condition that affects approximately 1 per 10,000 person/year [1, 2]. In uveitis, the incidence of retinal detachment is higher, ranging from 2 to 7% [3–6], and including the exudative, tractional and rhegmatogenous forms, the latter two requiring most frequently a surgical approach [7]. Nonetheless, the choice of the most suitable surgical technique remains unvalidated. Pars plana vitrectomy (PPV), scleral buckling (SB) or a combination of both have all been tried with variable outcomes [3, 5, 6, 8, 9]. Unlike the non-inflamed eye, the visual prognosis in uveitic RD was reported to be poor with variable anatomic and functional results [3–5]. The prognostic factors of RD surgery in the inflamed eye are poorly known. The purpose of this work was to study the functional and anatomical outcomes of surgery in rhegmatogenous RD in the context of intraocular inflammation (IOI) and to describe the prognostic factors of surgical failure.
Materials and methods
Design
This was a retrospective study of consecutive eyes undergoing vitreoretinal surgery for rhegmatogenous RD associated with IOI in two French tertiary care units (Pitié-Salpêtrière, Paris and Nantes University Hospital, Nantes, France) between 2012 and 2019. The study was carried out following the tenets of the Declaration of Helsinki and STROBE guidelines and approved by the institutional review board (IRB) of the French Society of Ophthalmology (00008855). This study does not contain any personal information that could lead to the identification of patients. Informed consent was obtained from all subjects. Consecutive patients (age ≥18 years) with a history of IOI and undergoing surgery for rhegmatogenous RD were included. The duration of RD had to be <12 months, and the post-operative follow-up had to be longer than 6 months. Exclusion criteria were: exudative or tractional RD, RD due to a macular hole, RD following post-surgical endophthalmitis or occurring within a year of any intraocular surgery.
Data collection
In Paris, a systematic review of medical records of patients with RD and IOI was conducted. In Nantes, a numerical platform linked to the clinical database of the hospital (eHOP, approved by the French data protection agency) was used to identify the patients. In total, data from 66 patients (71 eyes) was collected (Supplementary eFig. 1).
Anterior segment inflammation and vitritis were graded according to the standardization of uveitis nomenclature (SUN) group criteria [10]. Active uveitis was also defined according to the SUN [10], including anterior chamber cells ≥0.5+ and/or vitritis ≥0.5+ and/or presence of active vasculitis and/or presence of active chorioretinal lesions. The known risk factors for developing rhegmatogenous RD were collected when available (prior surgery, myopia etc.) [11]. For each patient, clinical data including the best corrected visual acuity (BCVA, LogMAR), IOI status, RD type, and macular status (ON or OFF, macula OFF being defined as a clinically detached macula or by the presence of a retinal detachment involving the central 1 mm of the macula on optical coherence tomography (OCT)) was collected at baseline (RD diagnosis) and different timepoints when available (1, 6, 12, 24, 36, 48, 60, 72 and 84 months). Data regarding the surgical technique (SB, PPV, or both) and tamponade (air, gas or SO) was recorded. Complications including significant intraocular hypertension (needing oral treatment or surgery), retinal re-detachment and phthisis bulbi were collected. Regarding IOI aetiology, all patients underwent standardized investigations [12] including complete blood count, kidney and liver tests, serum angiotensin-converting enzyme level, serology for syphilis, human immunodeficiency virus (HIV), interferon-gamma release assay test, human leucocyte antigen-B27, A29 and B51, chest imaging and an internal medicine work-up. Anterior chamber tap was performed when acute retinal necrosis (ARN), progressive outer retinal necrosis (PORN), viral uveitis, toxoplasmosis, endogenous endophthalmitis or vitreoretinal lymphoma were suspected, and/or in the absence of response to conventional anti-inflammatory treatment. IOI was subdivided into 4 categories: infectious, “autoimmune” (i.e. associated with a well-identified systemic autoimmune disease such as sarcoidosis), tumoral (i.e. vitreoretinal lymphoma) and “idiopathic or undifferentiated” (when repeated clinical, biological and imaging investigations remained negative).
Statistical analyses
For statistical analyses, the R software was used (URL https://www.R-project.org/.). The BCVA was tested using a decimal scale (Monoyer chart) converted to LogMAR for calculations (1.7 LogMAR for counting finger, 2.3 LogMAR for hand motion, 2.4 LogMAR for light perception and 2.6 LogMAR for no light perception). Quantitative variables were described using mean and standard deviation. Categorical variables were described using absolute and relative frequencies. Chi-square and Fisher exact tests were used when appropriate. Student’s t test and Wilcoxon test were used for comparison of two quantitative variables when appropriate. The level of significance was set at p < 0.05.
Results
Patients’ characteristics (Tables 1 and 2)
Table 1.
Patients’ characteristics at baseline.
| Infectious | VRL | Associated with systemic autoimmune disease | Idiopathic or undifferenciated | Overall | |
|---|---|---|---|---|---|
| (N = 53 eyes) | (N = 2 eyes) | (N = 5 eyes) | (N = 11 eyes) | (N = 71 eyes) | |
| Gender | |||||
| Females (N,%) | 25 (47.2%) | 1 (50%) | 4 (80%) | 5 (45.5%) | 35 (49.3%) |
| Age at IOI diagnosis (years) | |||||
| Mean (SD) | 50.5 (18) | 62 (4.2) | 38.4 (20.0) | 50.5 (17.5) | 50 (17.9) |
| Median [Min,Max] | 57 (18,80) | 62 (59,65) | 46 (15,56) | 52 (14,71) | 56 (14,80) |
| Age at RD diagnosis (years) | |||||
| Mean (SD) | 51.4 (17.6) | 62.5 (5) | 47.0 (19.0) | 53 (16.3) | 51.6 (17.2) |
| Median [Min,Max] | 57 (18,81) | 62.5 (59,66) | 47 (18,64) | 53 (19,71) | 55 (18,81) |
| Interval between IOI diagnosis and RD occurrence (months) | |||||
| Mean (SD) | 14 (42.1) | 7.5 (7.8) | 105 (109) | 29 (32.5) | 22.5 (52) |
| Median [Min,Max] | 3 (0,288) | 7.5 (2,13) | 96 (12,288) | 16 (0,84) | 4 (0,288) |
| Follow-up duration (months) | |||||
| Mean (SD) | 33.9 (22.7) | 21 (21.2) | 42 (27.2) | 49.6 (24) | 36.6 (23.6) |
| Median [Min,Max] | 24 (6,84) | 21 (6,36) | 40 (12,84) | 42 (12,84) | 28 (6,84) |
| Laterality (N,%) | |||||
| Right eye (N,%) | 29 (54.7%) | 0 (0%) | 4 (80.0%) | 6 (54.5%) | 39 (54.5%) |
| IOI anatomic location (N,%) | |||||
| Anterior uveitis (N,%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Intermediate uveitis (N,%) | 2 (3.8%) | 1 (50%) | 2 (40%) | 1 (9.1%) | 6 (8.5%) |
| Posterior Uveitis (N,%) | 6 (11.3%) | 0 (0%) | 0 (0%) | 1 (9.1%) | 7 (9.8%) |
| Panuveitis (N,%) | 45 (84.9%) | 1 (50%) | 3 (60%) | 9 (81.8%) | 58 (81.7%) |
| Immunosuppression (N,%) | 18 (34%) | 1 (50%) | 2 (40%) | 2 (18.2%) | 23 (32.4%) |
| Past ocular surgery (cataract and/or vitrectomy) (N,%) | 12 (22.6%) | 1 (50%) | 1 (20%) | 6 (54.5%) | 20 (28.2%) |
| Pseudophakia at baseline (N,%) | 11 (20.8%) | 1 (50%) | 1 (20%) | 6 (54.5%) | 19 (26.8%) |
| Myopia (N,%) | 14 (26.4%) | 0 (0%) | 2 (40%) | 3 (27.3%) | 19 (26.8%) |
| Lattice degeneration (N,%) | 0 (0%) | 0 (0%) | 0 (0%) | 2 (18.2%) | 2 (2.8%) |
| RD characteristics (N,%) | |||||
| Symptoms evocative of retinal detachment before diagnosis (N,%) | 27 (50.9%) | 0 (0%) | 4 (80%) | 5 (45.5%) | 36 (50.7%) |
| Macula OFF at diagnosis (N,%) | 38 (71.7%) | 1 (50%) | 4 (80%) | 9 (81.8%) | 52 (73.2%) |
| Grade C proliferative vitreoretinopathy at baseline (N,%) | 14 (26.4%) | 0 (0%) | 1 (20%) | 4 (36.4%) | 19 (26.8%) |
| Retinal tear found before surgery (N,%) | 17 (32.1%) | 1 (50%) | 2 (40%) | 8 (72.7%) | 28 (39.4%) |
| Retinal tear found during surgery (N,%) | 42 (79.2%) | 1 (50%) | 5 (100%) | 9 (81.8%) | 57 (80.3%) |
| Active inflammation at RD occurrence (N,%) | 19 (35.8%) | 1 (50%) | 0 (0%) | 7 (63.6%) | 27 (38%) |
| BCVA before surgery (LogMAR) | |||||
| Mean (SD) | 1.4 (0.7) | 1.2 (1.6) | 1.7 (0.9) | 1.6 (0.8) | 1.5 (0.8) |
| Median [Min,Max] | 1.7 (0,2.6) | 1.2 (0,2.3) | 2.3 (0.1,2.4) | 1.7 (0,2.4) | 1.7 (0,2.6) |
| Countable BCVA before surgery (N,%) | 24 (45.3%) | 1 (50%) | 1 (20%) | 5 (45.5%) | 31 (43.7%) |
Countable BCVA is ≥20/400 Snellen equivalent.
N number of eyes, % percentage, VRL vitreoretinal lymphoma, IOI intraocular inflammation, RD retinal detachment, BCVA best corrected visual acuity, SD standard deviation, Min minimum, Max maximum.
Table 2.
Anatomic and functional outcomes at different post-operative times points according to IOI aetiology.
| N (%) | BCVA at baseline Mean ± SD (range) | BCVA at 1-month Mean ± SD (range) | p-value | BCVA at 6 months Mean ± SD (range) | p-value | BCVA at 12 months Mean ± SD (range) | p-value | |
|---|---|---|---|---|---|---|---|---|
| Overall | 71 (100%) | N = 71 | N = 71 | N = 70 | N = 63 | |||
| 1.47 ± 0.78 (0–2.6) | 1.18 ± 0.7 (0–2.4) | 0.002 | 1.14 ± 0.74 (−0.1–2.6) | 0.0001 | 1.14 ± 0.75 (−0.1–2.4) | 0.001 | ||
| Infectious causes | 53 (74.7%) | N = 53 | N = 53 | N = 52 | N = 46 | |||
| 1.4 ± 0.7 (0–2.6) | 1.2 ± 0.7 (0.3–2.4) | 0.02 | 1.1 ± 0.7 (0–2.4) | 0.003 | 1 ± 0.7 (0.1–2.4) | 0.02 | ||
| VZV (ARN) | 12 (17%) | 1.7 ± 0.7 (0.4–2.4) | 1.2 ± 0.7 (0.3–2.4) | 1.3 ± 0.6 (0.3–2.3) | 1.4 ± 0.7 (0.5–2.4) | |||
| Endogenous endophthalmitis | 9 (12.7%) | 2 ± 0.3 (1–2.4) | 1.3 ± 0.7 (0.3–2.3) | 2 ± 0.8 (0.1–2.4) | 1 ± 0.9 (0.1–2.4) | |||
| CMV | 9 (12.7%) | 0.9 ± 0.7 (0–2.3) | 0.9 ± 0.8 (0.3–2.3) | 0.5 ± 0.4 (0–1) | 0.5 ± 0.3 (0.1–0.8) | |||
| HSV-2 (ARN) | 8 (11.3%) | 1.6 ± 0.9 (0.4–2.4) | 1.1 ± 0.6 (0.5–2.3) | 1.2 ± 0.7 (0.4–2.3) | 1.6 ± 0.6 (0.8–2.3) | |||
| Toxoplasmosis | 5 (7%) | 1 ± 0.8 (0.4–2.6) | 2 ± 0.6 (0.7–2.3) | 1 ± 0.9 (0.4–2.3) | 1 ± 0.7 (0.2–1.7) | |||
| CMV + VZV (ARN or PORN) | 3 (4.2%) | 0.7 ± 0.3 (0.5–1) | 0.6 ± 0.3 (0.4–1) | 1.2 ± 0.6 (0.5–1.7) | 0.7 ± 0.5 (0.3–1) | |||
| HSV-1 (ARN) | 2 (2.8%) | 2.3 ± 0 (2.3–2.3) | 2 ± 0.4 (1.7–2.3) | 1.2 ± 0.8 (0.6–1.7) | 1.6 ± 1.1 (0.8–2.4) | |||
| Tuberculosis | 1 (1.4%) | 1 | 1 | 0.2 | 0.1 | |||
| Syphilis | 1 (1.4%) | 1 | 2.3 | 0.3 | – | |||
| Toxocariasis | 1 (1.4%) | 0.7 | 0.9 | 1 | 1 | |||
| Rickettsiosis | 1 (1.4%) | 2.3 | 1.7 | 1.7 | 1.7 | |||
| EBV | 1 (1.4%) | 0.7 | 0.6 | 0.4 | 0.2 | |||
| Associated with systemic autoimmune disease | 5 (7%) | N = 5 | N = 5 | N = 5 | N = 5 | |||
| 1.8 ± 1 (0.1–2.4) | 1 ± 0.7 (0.1–1.7) | 0.05 | 1 ± 0.9 (0.1–2.3) | 0.1 | 0.8 ± 0.9 (0.2–2.3) | 0.1 | ||
| Sarcoidosis | 2 (2.8%) | 2.4 ± 0.1 (2.3–2.4) | 1.2 ± 0.8 (0.6–1.7) | 1.3 ± 1.4 (0.3–2.3) | 1.3 ± 1.4 (0.3–2.3) | |||
| Multiple sclerosis | 1 (1.4%) | 2.3 | 1.7 | 1.3 | 0.5 | |||
| Crohn’s disease | 1 (1.4%) | 1.7 | 1 | 1 | 0.9 | |||
| Sympathetic ophthalmia | 1 (1.4%) | 0.1 | 0.1 | 0.1 | 0.2 | |||
| VRL | 2 (2.8%) | N = 2 | N = 2 | N = 2 | N = 1 | |||
| 1.2 ± 1.6 (0–2.3) | 0.9 ± 1.1 (0.1–1.7) | – | 1.3 ± 1.8 (0–2.6) | – | – | |||
| Primary ocular | 1 (1.4%) | 0 | 0.1 | 0 | 0 | |||
| Extranodal NK/T | 1 (1.4%) | 2.3 | 1.7 | 2.6 | – | |||
| Idiopathic or undifferentiated | 11 (15.5%) | N = 11 | N = 11 | N = 11 | N = 11 | |||
| 1.6 ± 0.8 (0–2.4) | 1.3 ± 0.7 (0–2.3) | 0.3 | 1.2 ± 0.8 (−0.1–2.3) | 0.07 | 1.2 ± 0.7 (−0.1–2.4) | 0.08 |
| BCVA at 24 months Mean ± SD (range) | p-value | BCVA at 36 months Mean ± SD (range) | p-value | BCVA at the last follow-up Mean ± SD (range) | p-value | Retinal reattachment rate at the last follow-up | |
|---|---|---|---|---|---|---|---|
| Overall | N = 54 | N = 37 | N = 71 | N = 71 | |||
| 1.07 ± 0.8 (−0.1–2.4) | 0.002 | 1.08 ± 0.9 (−0.1–2.6) | 0.06 | 1.1 ± 0.88 (−0.1–2.6) | 0.001 | 64/71 (90.1%) | |
| Infectious causes | N = 39 | N = 24 | N = 53 | ||||
| 1 ± 0.8 (0.1–2.4) | 0.04 | 1 ± 0.9 (0.1–2.4) | 0.2 | 1.1 ± 0.9 (0–2.6) | 0.01 | 47/53 (89%) | |
| VZV (ARN) | 1.3 ± 0.8 (0.3–2.4) | 1.9 ± 0.8 (0.5–2.4) | 1.4 ± 0.8 (0.3–2.6) | 10/12 (83%) | |||
| Endogenous endophthalmitis | 1 ± 0.8 (0.1–2.4) | 1 ± 1 (0.1–2.4) | 1 ± 0.9 (0.1–2.4) | 8/9 (89%) | |||
| CMV | 0.6 ± 0.4 (0.2–1.1) | 0.2 ± 0.2 (0.1–0.4) | 0.4 ± 0.4 (0–1.1) | 9/9 (100%) | |||
| HSV-2 (ARN) | 1.7 ± 1 (0.3–2.4) | 1.2 ± 1 (0.3–2.3) | 1.6 ± 0.9 (0.3–2.4) | 6/8 (75%) | |||
| Toxoplasmosis | 0.8 ± 0.5 (0.1–1.3) | 0.7 ± 0.5 (0.2–1) | 0.4 ± 0.2 (0.1–0.7) | 5/5 (100%) | |||
| CMV + VZV (ARN or PORN) | 0.7 ± 0.5 (0.3–1) | 1 | 0.9 ± 0.7 (0.3–1.7) | 3/3 (100%) | |||
| HSV-1 (ARN) | – | – | 1.6 ± 1.1 (0.8–2.4) | 2/2 (100%) | |||
| Tuberculosis | 0.1 | 0.1 | 0.1 | 1/1 (100%) | |||
| Syphilis | – | – | 0.3 | 1/1 (100%) | |||
| Toxocariasis | 2.4 | – | 2.4 | 0/1 (0%) | |||
| Rickettsiosis | 1.7 | – | 1.7 | 1/1 (100%) | |||
| EBV | 1.7 | 2.3 | 2.3 | 1/1 (100%) | |||
| Associated with systemic autoimmune disease | N = 4 | N = 3 | N = 5 | ||||
| 0.8 ± 1 (0.1–2.3) | 0.2 | 1.1 ± 1 (0.2–2.6) | 0.5 | 0.8 ± 1 (0.1–2.6) | 0.1 | 5/5 (100%) | |
| Sarcoidosis | 1.2 ± 1.6 (0.1–2.3) | 2.6 | 1.4 ± 1.8 (0.1–2.6) | 2/2 (100%) | |||
| Multiple sclerosis | 0.4 | 0.5 | 0.5 | 1/1 (100%) | |||
| Crohn’s disease | – | – | 0.9 | 1/1 (100%) | |||
| Sympathetic ophthalmia | 0.2 | 0.2 | 0.1 | 1/1 (100%) | |||
| VRL | N = 1 | N = 1 | N = 2 | ||||
| – | – | 1.4 ± 1.8 (0.1–2.6) | 2/2 (100%) | ||||
| Primary ocular | 0.1 | 0.1 | 0.1 | 1/1 (100%) | |||
| Extra nodal NK/T | – | – | 2.6 | 1/1 (100%) | |||
| Idiopathic or undifferentiated | N = 10 | N = 9 | N = 11 | ||||
| 1.1 ± 0.8 (−0.1–2.4) | 0.03 | 1 ± 0.8 (−0.1–2.6) | 0.2 | 1.2 ± 0.8 (−0.1–2.6) | 0.1 | 10/11 (90.9%) |
Statistically significant p-values are in bold.
P-values represent the comparison between a specific time-point and baseline, no calculated p-value for VRL because of the limited number of included eyes. The last follow-up was after a mean duration of 36.6 ± 23.5 months (range: 6–84).
IOI intraocular inflammation, BCVA best corrected visual acuity, N number of eyes, SD standard deviation, VRL vitreoretinal lymphoma, VZV Varicella Zoster Virus, ARN Acute Retinal Necrosis, HSV Herpes Simplex Virus, CMV Cytomegalovirus, PORN Progressive Outer Retinal Necrosis, EBV Epstein-Barr Virus.
Patients’ characteristics are presented in Table 1. Briefly, there were 49.3% of females. The mean age at RD diagnosis was 50.0 ± 17.9 years and the mean follow-up was 36.6 ± 23.6 months. The mean duration of IOI prior to RD was 22.5 ± 52 months. At baseline, retinal detachment was unilateral in all but 4 patients (all of the latter had an infectious uveitis). Taking into account the whole cohort (Table 2), viral infection was the leading cause of IOI (35/71 eyes, 49.3%), including 65.7% of eyes with herpetic ARN (23/35 eyes), 25.7% with cytomegalovirus (CMV) retinitis (9/35 eyes), 5.7% with PORN (simultaneous CMV and varicella zoster (VZV) infection) (2/35 eyes) and 2.9% with an Epstein-Barr virus (EBV)- associated panuveitis (1/35 eyes). Approximately 13% (9/71) of eyes were diagnosed with an endogenous endophthalmitis secondary to Candida albicans in 3 eyes and to Streptococcus, Listeria, Pseudomonas and Klebsiella in 1 eye each. Thirty-two percent of RD cases (23/71) occurred in immunosuppressed individuals, including 52.2% with a human immunodeficiency virus (HIV) infection (12/23), 21.7% treated with corticosteroids and immunosuppressive therapy for uveitis (5/23), 21.7% treated with immunosuppressants for an organ transplant (5/23) and 4.3% treated with splenectomy (1/23). In immunosuppressed patients, IOI was of an infectious origin in 78.2% of cases (18/23). Patients’ characteristics were well balanced between the two centres (Paris and Nantes, Supplementary eTable 1, Supplements).
Characteristics of RD and IOI at the time of RD diagnosis (Table 1, Supplementary eFig. 2)
All retinal tears were peripheral. At diagnosis, grade C proliferative vitreoretinopathy (PVR) was present in 26.8% of eyes (19/71), 73.7% (14/19) of which had an infectious IOI. The median interval between the diagnosis of IOI and RD was 3 versus 20 months in infectious and non-infectious uveitis respectively (p = 0.003) (Supplementary eFig. 2, Supplement). More specifically, RD occurred within the first year of IOI in 89% of eyes with infectious versus 38% of those with non-infectious uveitis. IOI was active in only 36.6% of eyes (26/71) at RD diagnosis. Among eyes with active uveitis at baseline, 61.5% (16/26) had an infectious cause for their IOI (75.1% viral, 12.5% endogenous endophthalmitis, 6.2% toxoplasmosis and 6.2% toxocariasis).
Visual outcomes - quantitative outcomes (Table 2, Supplementary eFig. 3)
Taking into account the whole cohort (Table 2, Supplementary eFig. 3), the mean BCVA was 1.47 ± 0.78 LogMAR at baseline improving to 1.14 ± 0.75 LogMAR at 12 months (p = 0.001) and to 1.1 ± 0.9 LogMAR at the last follow up (p = 0.001) (i.e. after a mean duration of 36.6 ± 23.6 months). The evolution of BCVA depending on IOI aetiology is presented in more details in Table 2.
Regarding the location of IOI, eyes with panuveitis (58/71 eyes, 81.7%) had a mean baseline BCVA of 1.54 ± 0.7 LogMAR, improving to 1.27 ± 0.75 LogMAR at 12 months (p = 0.01) and to 1.2 ± 0.9 LogMAR at the last follow-up (p = 0.005), after a mean duration of 36.1 ± 24.7 months. Eyes with intermediate or posterior uveitis (13/71 eyes, 18.3%) had a mean baseline BCVA of 1.14 ± 0.95 LogMAR (p = 0.2 versus panuveitis) improving to 0.6 ± 0.51 LogMAR at 12 months (p = 0.02 versus baseline and p = 0.001 versus panuveitis) and to 0.65 ± 0.74 LogMAR at the last follow-up, after a mean duration of 38.8 ± 18.8 months (p = 0.1 versus baseline and p = 0.03 versus panuveitis at the last follow-up).
Regarding RD characteristics, the mean baseline BCVA was respectively 0.8 ± 0.7 LogMAR (range:0-2.4) and 1.7 ± 0.6 LogMAR (range: 0.4–2.6) in macula ON (19/71, 26.8%) and OFF (52/71, 73.2%) eyes (p = 0.001). At 12 months of surgery, the mean BCVA was 0.93 ± 0.86 LogMAR (range: −0.1−2.4) and 1.2 ± 0.71 LogMAR (range: 0.1–2.4) in macula ON and OFF eyes respectively (p = 0.3). At the last follow-up (average time of 36.6 ± 23.6 months), the mean BCVA was still comparable between macula-ON and OFF eyes (0.8 ± 0.9 LogMAR versus 1.2 ± 0.8 LogMAR, p = 0.1). However, while the pre-operative and last follow-up BCVA figures were comparable in macula ON eyes (p = 0.9), macula-OFF eyes showed a significant BCVA improvement after surgery (p < 0.001, last follow-up versus baseline).
Regarding the perioperative management, 26.7% of eyes (19/71) received steroids according to the following regime: intravenous methylprednisolone (500 mg/day) given 3 days prior to surgery; perioperative intravenous dexamethasone (4 mg/kg); and post operative oral steroids (dose and duration depending on uveitis aetiology). At the last follow-up, BCVA was not statistically different between eyes that received perioperative steroids according to the aforementioned regime and those that did not (p = 0.2).
The use of 360° laser retinopexy was not associated with better BCVA outcomes at the 12-month (1.2 ± 0.7 LogMAR versus 0.9 ± 0.7 LogMAR in eyes receiving versus those not receiving 360° laser (p = 0.3)) nor the last follow-up (1.2 ± 0.9 LogMAR versus 0.8 ± 0.7 LogMAR in eyes with versus without 360-laser respectively, p = 0.04).
Visual outcomes- qualitative outcomes and prognostic factors of visual recovery (Table 3)
Table 3.
Univariate and multivariate analyses of the predictive factors for having a countable BCVA at 12 months.
| Univariate model | Multivariate model | |||||
|---|---|---|---|---|---|---|
| OR | IC95% | p-value | OR | IC95% | p-value | |
| Countable BCVA before surgery | 3.70 | [1.33–11.96] | 0.017 | 3.4 | [0.09–2.47] | 0.04 |
| Countable BCVA at 1 month of surgery | 8.74 | [2.99–28.09] | <0.001 | |||
| Countable BCVA at 6 months of surgery | 41 | [11.01–195.36] | <0.001 | |||
| Macula ON RD | 1.75 | [0.57–6.09] | 0.35 | |||
| Active intraocular inflammation at RD occurrence | 0.62 | [0.22–1.69] | 0.34 | |||
| Use of cryotherapy during first surgery | 2.12 | [0.79–5.94] | 0.14 | |||
| Retinal tear found during first surgery | 3.53 | [0.9–15.36] | 0.07 | |||
| PVR grade C at baseline | 0.46 | [0.15–1.38] | 0.16 | |||
| RD relapse at any time of the follow-up | 0.59 | [0.20–1.80] | 0.35 | |||
| Gaz tamponade | 6.55 | [1.12–124.80] | 0.08 | |||
Statistically significant p-values are in bold.
Countable BCVA is ≥20/400 Snellen equivalent.
OR odds ratio, IC95% 95% confidence interval, BCVA best corrected visual acuity, RD retinal detachment, PVR proliferative vitreo retinopathy.
At baseline 43.7% of eyes (31/71) had a countable visual acuity (ie. BCVA ≥ 20/400 Snellen) improving to 63.5% (40/63) at 12 months (p = 0.008) and to 66.2% (47/71) at the last follow-up (p = 0.006). Among the 33.8% of eyes with an uncountable BCVA at the last follow-up, 17% (N = 4) had no light perception. The factors predicting the event “having a countable BCVA at 12 months” are shown in Table 3. On univariate analysis, having a countable BCVA at baseline, 1 month and 6 months of surgery was associated with the event “having a countable BCVA at 12 months” (all p < 0.05). On multivariate analysis, a “countable BCVA at baseline” was the only predictive factor of this functional outcome.
Surgical characteristics
RD surgery was performed after a median interval of 4 days (mean:11.5 days, range: 0–240), mainly under general anaesthesia (98.6% of cases). The delay between RD and surgery was similar among all uveitis aetiologies and macula-ON and OFF eyes. The first surgery consisted in PPV alone in 87.3% of cases (62/71), SB alone in 4.2% of cases (3/71), and PPV + SB in 8.4% of eyes (6/71). Sixty-seven eyes received tamponade, including 80.3% (57/71) with silicone oil (SO) (1000 cSt: 98.3%, 3800 cSt: 1.7%), and 14.1% (10/71) with gas (SF6: 20%, C2F6: 40%, C3F8: 40%). Silicone oil was removed in 64.9% of eyes (37/57) after a mean duration of 9.2 ± 5.5 months. Sixty-six percent of eyes (45/68) undergoing vitrectomy received 360° laser retinopexy while 27.9% (19/68) received focal laser at the site of the retinal tear and 4 eyes (5.9%) received cryotherapy alone. At the time of RD diagnosis, 73.2% of eyes (52/71) were phakic. Among them, 19.2% (10/52) underwent simultaneous cataract and RD surgery. Among eyes being still phakic after the first RD surgery, 90.5% (38/42) developed a cataract during the follow-up and 89.5% (34/38) of them underwent cataract surgery after a mean interval of 8.9 ± 6.5 months.
Retinal reattachment and its prognostic factors (Tables 2 and 4, Fig. 1)
Table 4.
Uni- and multivariate analyses of the predictive factors of retinal reattachment at 12 months.
| Univariate model | Multivariate model | |||||
|---|---|---|---|---|---|---|
| OR | IC95% | p-value | OR | IC95% | p-value | |
| Countable BCVA before surgery | 3.33 | [0.46–67.09] | 0.2 | |||
| Macula ON | 1.5 | [0.20–30.40] | 0.7 | |||
| PVR grade C at baseline | 0.57 | [0.09–4.58] | 0.5 | |||
| Retinal tear found during first surgery | 4.5 | [0.53–31.5] | 0.12 | |||
| RD relapse within the 6 first weeks of surgery | 0.08 | [0.01–0.55] | 0.01 | 0.03 | [−7.35; −0.86] | 0.02 |
| Withdrawal of silicone oil during follow-up | 9.0 | [1.21–183] | 0.02 | |||
Statistically significant p-values are in bold.
Countable BCVA is ≥20/400 Snellen equivalent.
OR odds ratio, IC95% 95% confidence interval, BCVA best corrected visual acuity, RD retinal detachment, PVR proliferative vitreo retinopathy.
Fig. 1. Retinal reattachment overtime.
A Percentage of patients with retinal reattachment at the last visit depending on the number of surgeries. B Kaplan–Meier survival curve showing the time to relapse after initial retinal detachment surgery.
Retinal reattachment was achieved after one surgery in 74.6% of cases (Fig. 1A). Seventy-eight percent of recurrences (14/18) occurred during the first year (Fig. 1B). Of the 18 eyes with redetachment, 10 (55.5%) had an early recurrence (ie. during the 6 first weeks). No redetachment was noted in patients undergoing SB alone or SB associated with PPV. The redetachment rate in eyes undergoing PPV alone was 29% (18/62). This rate was 20% in eyes receiving gas tamponade compared to 28% in those receiving SO (p = 0.7).
At 12 months, the reattachment rate after one or multiple surgeries was 92.1% (58/63). At the last follow-up (after a mean period of 36.6 ± 23.6 months), the reattachment rate after one or multiple surgeries was 90.1% (64/71). This corresponded to a mean surgery number of 1.4 ± 0.8 (range: 1–5). This final rate did not differ between the infectious and non-infectious causes of uveitis (88% versus 94% at the last follow-up, p = 0.8) (Table 2), nor was it different between macula-ON and OFF eyes (94% versus 88.5%, p = 0.7); or in eyes with intermediate and posterior uveitis (92.3%) versus those with panuveitis (89.7%) (p = 1). Eyes with active versus inactive IOI at baseline had a final reattachment rate of 85% versus 93% (p = 0.4). There was no significant association between the surgical delay and the reattachment rate at 12 months (univariate analysis, reference = delay ≥ 4 days, OR = 0.67, [0.08–4.27], p = 0.67).
Using a univariate model (Table 4), the absence of RD recurrence within the first 6 weeks of surgery (OR = 12.5, p = 0.01) and silicone oil withdrawal (OR = 9, p = 0.02) were both associated with a higher likelihood of retinal reattachment at 12 months. On multivariate analysis, the only predictive factor of retinal reattachment at 12 months was the absence of an early RD recurrence (≤6 weeks).
Safety
During the follow-up, 28% of eyes (16/57) with SO tamponade developed significant intraocular hypertension (see methods). Phthisis bulbi occurred in two eyes. No postoperative endophthalmitis was observed.
Discussion
Key results and interpretation
This is the largest cohort investigating the outcomes of vitreoretinal surgery in rhegmatogenous RD following IOI. The characteristics of included eyes matched those of previous publications. We found similar demographic characteristics compared to the literature [3–6]. Regarding the anatomic type of uveitis, 91.5% of eyes had either posterior uveitis or pan-uveitis, similar to previous studies [3, 4, 6]. We found 74.7% of infectious causes, the majority of which being triggered by ARN or CMV retinitis, corresponding to the upper range (52–70%) described in other publications [3–6]. Interestingly, we reported on the first RD cases complicating ocular rickettsiosis and Crohn’s disease, and the third case of RD associated with EBV-uveitis [13]. The percentage of eyes with baseline BCVA ≤ 20/400 was 56.3% and 71.8% when using a BCVA cut-off of 20/200, comparable to the 53.4–82.8% rate reported in other publications [3, 5, 6]. Seventy-three percent of eyes were macula-OFF, which is higher than in primary rhegmatogenous RD (50.5%) [14], but in the same range as in RD cases complicating ARN [8]. This may be explained by the decreased awareness of visual loss in uveitic patients, especially in those with chronic vitritis or macular oedema. This may also explain the surgical delay that showed a large range in our series. Slingsby et al. showed that noninfectious uveitis had a higher risk of PVR compared to primary rhegmatogenous RD [15]. In this study, the rate of PVR was comparable among all IOI aetiologies, suggesting that IOI can promote PVR to a similar degree regardless of its cause [15]. Regarding RD predisposing factors, we found that only 26.8% of eyes were myopic, which is lower than the 32.2–47.3% rate reported in primary rhegmatogenous RD [2, 16] or the 46% rate found in uveitic RD [6]. The presence of lattice degeneration was poorly reported on our patients’ charts which did not allow us to include this factor in our analyses. Lastly, we did not include eyes undergoing intraocular surgery≤1 year of RD diagnosis in order to limit any bias due to iatrogenically caused retinal detachments.
The mean interval between uveitis onset and RD was slightly inferior than in previous studies: 22.5 months versus 24–34 months reported elsewhere [4, 6]. We showed that RD occurred significantly earlier in infectious versus non-infectious IOI, likely due to the earlier occurrence and/or higher frequency of retinal breaks in viral retinitis, the main cause of RD in our cohort. In fact, RD secondary to infectious aetiologies occurred after a median duration of 3 months (versus 20 months in non-infectious causes), with 89% of detachments occurring within a year of uveitis onset. It is interesting to note that almost 50% of patients with infectious uveitis reported no symptom evocative of RD at the time of its diagnosis. These two observations suggest the importance of a regular and systematic follow-up.
We found a similar proportion of eyes improving their BCVA after surgery at the one-year follow-up (55.6%) than in Kunavisarut et al. (52.6%) [3]. This corresponded to a 63.5% rate of countable BCVAs. This rate did not differ between the infectious, autoimmune, lymphomatous or idiopathic causes of IOI. Compared to previous reports, the reattachment rate after one surgery was 74.6% versus 41–52% in uveitic RD [3–6] and 76–94% in non-uveitic RD [17, 18]. In our study, this reattachment rate did not differ depending on IOI aetiology. The final reattachment rate (≥1 surgery) was 90.1%, which was lower than in non-uveitic RD (98.1–100% [14]), but higher than other studies reporting on RD following IOI (73.7–88%) [3–6]. The cause for this better anatomical outcome could be the use of improved surgical techniques, the higher rate of SO tamponade (80.3% versus 65% in Kunavisarut et al.) [3], or the higher rate of 360 laser retinopexy (66.2% in our study, not reported elsewhere). This rate might also be influenced by the absence of early SO removal and the long-term tamponade (mean: 9.2 ± 5.5months). Another explanation could be the use of aggressive peri-operative anti-inflammatory treatments in our study, allowing better controlled inflammation at the time of surgery. It should be noted that infectious RD (and more specifically ARN) is often associated with small-size retinal holes, which may explain the lower anatomical success rate (65.2% after one surgery, data not shown) in this category of eyes. On multivariate analysis, RD recurrence within the first six weeks of surgery was the only predictive factor for anatomical failure at the 12-month follow-up. While not surprising, this previously unreported finding is useful because it provides surgeons with a timeline that may help them adapt their monitoring and the information they share with their patients. While the number of eyes undergoing SB was limited (N = 9), the use of an external approach yielded the best anatomical results (0% re-detached). Echegaray et al. [19] and more recently Moharana et al. [8] also suggested better anatomical outcomes when SB was combined with vitrectomy in non-uveitic rhegmatogenous RD and in RD associated with ARN. An external approach may in fact alleviate the indissociable peripheral vitreous tractions that are often seen in inflamed eyes.
Regarding the predictive factors of visual recovery, univariate and multivariate analyses showed that having a countable preoperative BCVA was associated with the event “having a countable BCVA at 12 months”. The association between a poor visual outcome and a baseline BCVA<counting fingers was previously reported by Kunavisarut et al. in uveitic RD [3] and has long been known in non-uveitic RD [20].
Limitations
There are limitations to this study that should be mentioned. First, they include its retrospective nature and the small number of inclusions due to the rarity of RD in IOI. On the other hand, this series is one of the largest ever published on this subject. While we cannot retrospectively check for small variations in surgical techniques, we believe that they are slight enough not to prevail in the final outcome. In fact, all surgeries were performed by physicians specialized in both vitreoretinal surgery and IOI. Meanwhile, including patients from one surgeon only would have limited even more the external validity of the study.
Although eyes undergoing SB, either alone or in combination with vitrectomy, had the best anatomical outcomes, both the small sample size and missing data prevented us from integrating the surgical technique into the multivariate model. Prospective randomized studies are underway to elucidate the role of SB as an adjunct to vitrectomy in uveitic RD. In the meantime, we believe that SB may be beneficial in combination with PPV in selected cases, allowing to alleviate vitreous adherences that are sometimes located very posteriorly in eyes with IOI.
Regarding the functional outcomes, data regarding specific macular characteristics (e.g. macula oedema) or the presence of a posterior capsule opacification were not available for all patients, inducing a possible bias that could not be taken into account in the final statistical models.
Conclusion
Rhegmatogenous retinal detachment associated with IOI bares a generally favourable anatomical prognosis. Baseline BCVA of ≥20/400 and absence of retinal redetachment during the first 6 weeks of follow-up were associated with improved VA and surgical success at 1 year. No redetachment was seen in cases receiving scleral buckling. These observations may prove useful in helping physicians with the prediction of their patients’ prognosis. These results suggest a potential adjunctive role of scleral buckling in addition to vitrectomy surgery for the treatment of RD cases with IOI.
Summary
What was known before
The outcomes of retinal detachment (RD) surgery in inflamed eyes are poor, and no prognostic factors are known to date.
What this study adds
Baseline BCVA over 20/400 and absence of RD relapse during the first 6 weeks were associated with the best results.
The best surgical outcomes (100% of reattachment) were found when scleral buckling (SB) was used alone or in association with par plana vitrectomy (PPV).
These results provide easily available figures to predict the prognosis. They also question the validity of adding SB to PPV in selected cases.
Supplementary information
Author contributions
Conceptualization: Sara Touhami, Bahram Bodaghi. Methodology: Sara Touhami. Validation: Bahram Bodaghi, Michel Weber, Sara Touhami. Investigation: Adam Mainguy. Data Curation: Adam Mainguy. Christine Fardeau, Adélaïde Toutée, Olivier Lebreton, Hélène Massé. Writing – Original Draft Preparation: Adam Mainguy, Sara Touhami. Writing – Review & Editing: Adam Mainguy, Sara Touhami, Michel Weber. Supervision: Sara Touhami.
Data availability
All data generated or analysed during this study are included in this published article.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
The online version contains supplementary material available at 10.1038/s41433-024-03300-0.
References
- 1.Steel D. Retinal detachment. BMJ Clin Evid. 2014;2014:0710. [PMC free article] [PubMed]
- 2.Mitry D, Charteris DG, Fleck BW, Campbell H, Singh J. The epidemiology of rhegmatogenous retinal detachment: geographical variation and clinical associations. Br J Ophthalmol. 2010;94:678–84. [DOI] [PubMed] [Google Scholar]
- 3.Kunavisarut P, Srisomboon T, Patikulsila D, Choovuthayakorn J, Watanachai N, Chaikitmongkol V, et al. Risk Factors for Development of Rhegmatogenous Retinal Detachment in Patients with Uveitis. Ocul Immunol Inflamm. 2019;27:681–5. [DOI] [PubMed] [Google Scholar]
- 4.De Hoog J, Ten Berge JC, Groen F, Rothova A. Rhegmatogenous retinal detachment in uveitis. J Ophthalmic Inflamm Infect. 2017;7:22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Arevalo JF, Lasave AF, Al Ghamdi TA, Gupta V, Kozak I, Al Dhibi HA, et al. Retinal detachment and uveitis at a tertiary center over 10 years: the King Khaled Eye Specialist Hospital (KKESH) Uveitis Survey Study Group. Graefes Arch. Clin. Exp. Ophthalmol. Albrecht Von Graefes Arch Klin Exp Ophthalmol. 2019;257:1857–61. [DOI] [PubMed] [Google Scholar]
- 6.Kerkhoff FT, Lamberts QJ, van den Biesen PR, Rothova A. Rhegmatogenous retinal detachment and uveitis. Ophthalmology. 2003;110:427–31. [DOI] [PubMed] [Google Scholar]
- 7.Ganesan S, Rishi E. Surgical implications in exudative retinal detachment. Sci J Med Vis Res. 2017.
- 8.Moharana B, Dogra M, Singh SR, Ravikumar B, Tigari B, Katoch D, et al. Outcomes of 25-gauge pars plana vitrectomy with encircling scleral band for acute retinal necrosis-related rhegmatogenous retinal detachment. Indian J Ophthalmol. 2021;69:635–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Young-Zvandasara T, Brunner D, Welch S, Sims JL, De Souza C, Cunningham WJ, et al. Rhegmatogenous retinal detachment presentation and surgery in uveitic eyes. Br J Ophthalmol. 2023;107:116–20. [DOI] [PubMed] [Google Scholar]
- 10.Jabs DA, Nussenblatt RB, Rosenbaum JT. Standardization of Uveitis Nomenclature (SUN) Working Group. Standardization of uveitis nomenclature for reporting clinical data. Results of the First International Workshop. Am J Ophthalmol. 2005;140:509–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ghazi NG, Green WR. Pathology and pathogenesis of retinal detachment. Eye Lond Engl. 2002;16:411–21. [DOI] [PubMed] [Google Scholar]
- 12.Sève P, Cacoub P, Bodaghi B, Trad S, Sellam J, Bellocq D, et al. Uveitis: Diagnostic work-up. A literature review and recommendations from an expert committee. Autoimmun Rev. 2017;16:1254–64. [DOI] [PubMed] [Google Scholar]
- 13.Chan EW, Sun V, Eldeeb M, Kapusta MA. Epstein-barr virus acute retinal necrosis in an immunocompetent host. Retin Cases Brief Rep. 2021;15:412–6. [DOI] [PubMed] [Google Scholar]
- 14.Haugstad M, Moosmayer S, Bragadόttir R. Primary rhegmatogenous retinal detachment - surgical methods and anatomical outcome. Acta Ophthalmol. 2017;95:247–51. [DOI] [PubMed] [Google Scholar]
- 15.Slingsby TJ, Pecen PE, Palestine AG. Outcomes and Complications Associated with Noninfectious Uveitis in Patients Presenting with Rhegmatogenous Retinal Detachment. Ophthalmol Retin. 2020;4:823–8. [DOI] [PubMed] [Google Scholar]
- 16.Ashrafzadeh MT, Schepens CL, Elzeneiny II, Moura R, Morse P, Kraushar MF. Aphakic and phakic retinal detachment. I. Preoperative Findings. Arch Ophthalmol. 1973;89:476–83. [DOI] [PubMed] [Google Scholar]
- 17.Sharma YR, Karunanithi S, Azad RV, Vohra R, Pal N, Singh DV, et al. Functional and anatomic outcome of scleral buckling versus primary vitrectomy in pseudophakic retinal detachment. Acta Ophthalmol Scand. 2005;83:293–7. [DOI] [PubMed] [Google Scholar]
- 18.Brazitikos PD, Androudi S, Christen WG, Stangos NT. Primary pars plana vitrectomy versus scleral buckle surgery for the treatment of pseudophakic retinal detachment: a randomized clinical trial. Retin Phila Pa. 2005;25:957–64. [DOI] [PubMed] [Google Scholar]
- 19.Echegaray JJ, Vanner EA, Zhang L, Fortun JA, Albini TA, Berrocal AM, et al. Outcomes of Pars Plana Vitrectomy Alone versus Combined Scleral Buckling plus Pars Plana Vitrectomy for Primary Retinal Detachment. Ophthalmol. Retina 2021;5:169–75. [DOI] [PubMed] [Google Scholar]
- 20.Burton TC. Recovery of visual acuity after retinal detachment involving the macula. Trans Am Ophthalmol Soc. 1982;80:475–97. [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data generated or analysed during this study are included in this published article.