To evaluate the efficacy and safety of prompt intravitreal triamcinolone acetonide injection (4 mg/0.1 mL) vs intravitreal bevacizumab injection (1.25 mg/0.05 mL) compared with observation in the management of extensive exudative retinal detachment secondary to posterior uveal melanoma.
Retrospective, nonrandomized, interventional case series.
setting: Institutional. patients: Ninety-six patients affected by posterior uveal melanoma with large exudative retinal detachment (>10 mm in largest basal diameter) were included. intervention: Patients received intravitreal triamcinolone acetonide (32 eyes) or intravitreal bevacizumab (32 eyes) at plaque removal. Thirty-two patients served as controls (observation group). All groups were matched for age, sex, initial tumor thickness and largest basal diameter, largest exudative retinal detachment basal diameter, tumor location, and Bruch membrane rupture. Patients underwent monthly follow-up examinations in the first 6 months and every 3 months thereafter. Follow-up was longer than 24 months. main outcome measure: Exudative retinal detachment resolution (B-scan ultrasonography), steroid-induced cataract, steroid-induced increased IOP.
Follow-up was 37 ± 7 months. Marked exudative retinal detachment regression was documented in 22 (69%) intravitreal triamcinolone acetonide–treated vs 11 (34%) intravitreal bevacizumab–treated and 9 (28%) untreated eyes ( P = .0007 and P = .0001, respectively). No statistical significance was found between intravitreal bevacizumab group vs observation group ( P = .45) Steroid-induced cataract was observed in 4 intravitreal triamcinolone acetonide–treated patients (12%). Neither steroid-induced increased IOP nor other short- or long-term side effects were documented.
Intraoperative intravitreal triamcinolone acetonide injection induces earlier and marked exudative retinal detachment resolution after brachytherapy of posterior uveal melanoma. Risk and benefit should be balanced vs steroid-induced cataract.
Eye-preserving radiation treatment has emerged as the standard of care for the management of most uveal melanoma since the Collaborative Ocular Melanoma Study confirmed equivalent life prognosis following brachytherapy compared with enucleation in medium-sized tumors. Although the goal of treating the tumor and preserving the eye is achieved in most cases, the preservation of vision remains unpredictable. This is mainly attributable, at least in medium-to-large tumors, to the persistence (or increase) after brachytherapy of the uveal melanoma–related exudative retinal detachment, with its long-term side effects (mainly retina ischemia and neovascular glaucoma). Although exudative retinal detachment can be caused by eye injury, inflammation, or vascular abnormalities, it is clinically detected in up to 75% of eyes with uveal melanoma, and it is considered the most common retinal side effect accompanying this tumor. Exudative retinal detachment is characterized by subretinal fluid accumulation secondary to fluid homeostasis alteration across the inner and outer blood-retinal barriers. Enhanced production of vascular endothelial growth factor (VEGF), nitric oxide, oxidative stress, and inflammation are considered the main factors that increase the permeability of both the inner and outer blood-retinal barriers.
Several proinflammatory cytokines are significantly expressed in uveal melanoma eyes. Ly and associates recently suggested that the presence of uveal melanoma leads to the production of a proinflammatory cytokine pattern that promotes tumor survival by creating a new microenvironment. Evidence of subclinical inflammation in uveal melanoma eyes was also previously reported by in vivo laser photometry measurements, detecting increased anterior chamber flare in untreated eyes with uveal melanoma. An increased flare in uveal melanoma eyes was also strongly correlated with the presence of uveal melanoma–related exudative retinal detachment.
One of the major group of proinflammatory cytokines is the VEGF family. VEGF-A is secreted by various types of tumor cells and can cause increased vascular permeability, endothelial cell growth, angiogenesis, monocyte activation, and chemotaxis. Uveal melanoma is associated with increased concentrations of VEGF-A in the aqueous humor. Recently, Missotten and associates suggested that VEGF-A concentration in the aqueous humor, exudative retinal detachment, and tumor size are interrelated factors.
A few years ago, we started a prospective pilot study to treat patients with extensive exudative retinal detachment secondary to uveal melanoma with intravitreal bevacizumab or intravitreal triamcinolone acetonide, mainly because of the lack of other truly safe options (see discussion).
The aim of this study is to retrospectively evaluate the efficacy and safety of prompt intravitreal triamcinolone acetonide vs intravitreal bevacizumab compared to observation in the management of exudative retinal detachment secondary to posterior uveal melanoma.
Materials and Methods
Setting and Design
This was an institutional, retrospective, nonrandomized, interventional case series. Patients were recruited from those referred to the Ophthalmic Oncology units of both the G.B. Bietti Foundation and the Department of Ophthalmology of the University of Padova. Treated patients were retrospectively recruited from those previously included in a pilot prospective study on safety of intravitreal triamcinolone acetonide and intravitreal bevacizumab in the management of exudative retinal detachment secondary to posterior uveal melanoma.
Patients were affected by posterior uveal melanoma with extensive exudative retinal detachment (>10 mm in largest basal diameter) and planned to undergo plaque brachytherapy (iodine-125 [I-125], dose of 85–100 Gy at tumor apex, dose rate of 0.60–1.05 Gy/h). Each patient underwent a baseline (preoperative) full ophthalmologic examination. The ophthalmoscopic aspect of the tumors was documented by fundus photography. Tumor and exudative retinal detachment dimensions were also documented using A- and B-scan ultrasonography. Clinical and demographic characteristics were collected, including age, sex, affected eye, tumor location (choroid vs ciliary body), tumor largest basal diameter (based on B-scan ultrasonography), tumor thickness (based on A-scan ultrasonography), exudative retinal detachment largest basal diameter (based on B-scan ultrasonography performed with patient lying on a stretcher), and presence of Bruch membrane rupture. Liver enzymes and liver ultrasonography were also used to evaluate the presence of metastatic disease at baseline. Ninety-six patients treated with I-125 brachytherapy, with a follow-up longer than 24 months, were included in this retrospective study.
Treated patients received 1 intravitreal triamcinolone acetonide injection at plaque removal (32 eyes) or a loading phase of 3 intravitreal bevacizumab injections (32 eyes), performed at 4-week intervals (the first intravitreal bevacizumab injection was performed during plaque removal). Thirty-two patients served as controls. All groups were matched for age, sex, initial tumor thickness and largest basal diameter, largest exudative retinal detachment basal diameter, tumor location, and Bruch membrane rupture. Clinical and demographic characteristics related to different treatment groups are reported in Table 1 . Cortico-responder patients were preliminarily excluded by the intravitreal triamcinolone acetonide group on the basis of a preoperative test.
|Intravitreal Triamcinolone Acetonide Group||Intravitreal Bevacizumab Group||Observation Group||All Groups|
|Age (y), mean (range)||56 ± 14 (33–84)||55 ± 15 (30–83)||58 ± 17 (37–86)||56 ± 15 (30–86)|
|Sex: male/female||14/18 (44%/56%)||15/17 (47%/53%)||17/15 (53%/47%)||46/50 (45%/55%)|
|Affected eye: right/left||18/14 (56%/44%)||19/13 (59%/41%)||21/11 (66%/34%)||58/38 (61%/39%)|
|Tumor location: choroidal/ciliary body||24/8 (75%/25%)||26/6 (81%19%)||25/7 (78%22%)||75/21 (78%/22%)|
|Baseline tumor dimensions|
|Thickness (mm)||6.7 ± 1.9||6.5 ± 1.9||6.7 ± 2.0||6.6 ± 1.9|
|(range, 4.0–10)||(range, 4.2–10)||(range, 4.0–10)||(range, 4.0–10)|
|Largest basal diameter (mm)||13.6 ± 3.3||12.9 ± 3.4||13.0 ± 3.5||13.3 ± 3.4|
|(range, 8–18)||(range, 7.6–18)||(range, 7.1–17)||(range, 7.1–18)|
|Baseline exudative retinal detachment dimension: largest basal diameter (mm)||15.4 ± 3.5||14.6 ± 3.3||15.5 ± 3.0||15.1 ± 3.5|
|(range, 10–21)||(range 10–20)||(range 10–19)||(range 10–21)|
|Bruch membrane rupture||14 (43.7%)||15 (46.8%)||12 (37.5%)||41 (43%)|
|Mean dose rate||0.85 ± 0.14 Gy/h||0.86 ± 0.14 Gy/h||0.86 ± 0.13 Gy/h||0.86 ± 0.14 Gy/h|
|Mean dose at tumor apex||91 ± 5 Gy||90 ± 4 Gy||91 ± 5 Gy||91 ± 5 Gy|
Sterile technique was used to prepare 4.0 mg/0.1 mL of triamcinolone acetonide from a single-use vial of 40 mg (Kenalog; Bristol-Myers-Squibb, Princeton, New Jersey, USA), immediately before the injection procedure. Bevacizumab injections (1.25 mg/0.05 mL) were prepared in sterile conditions by the hospital pharmacy from a single-use vial of 100 mg (Avastin; Genentech, Cambridge, Massachusetts, USA). Solutions were injected using a 27-gauge (intravitreal triamcinolone acetonide) or 30-gauge (intravitreal bevacizumab) needle via standard pars plana approach (3.5–4.0 mm posterior to the limbus), avoiding retinal quadrants occupied by tumor and/or exudative retinal detachment. Injection was performed at the time of I-125 plaque removal. The injection site was recorded. Fundus was observed for side effects immediately after injection using indirect ophthalmoscopy. Anterior chamber paracentesis was never performed. Patients underwent postoperative examination 1 and 7 days after injection.
Retreatment criteria were: (1) no or partial response after first injection (intravitreal triamcinolone acetonide) or loading phase (intravitreal bevacizumab); and (2) partial response after subsequent injections. Retreatment was performed at 3-month intervals. Treatment failure was finally managed by observation alone. Laser photocoagulation was employed to treat reattached ischemic retina in each patient having any response to treatment (partial or marked), or spontaneous resolution (partial or marked) in the observation group.
Main Outcome Measure
No response was defined as exudative retinal detachment largest basal diameter reduction of less than 20%. Partial response was defined as exudative retinal detachment largest basal diameter reduction between 20% and 80%. Marked response was defined as exudative retinal detachment largest basal diameter reduction of more than 80%. Treatment failure was defined as absence of marked response at the end of treatment. Steroid-induced increased IOP was defined as an IOP elevation to 24 mm Hg or higher. Steroid-induced cataract was assessed by slit-lamp examination after dilation of the pupil compared with baseline recorded data.
Follow-up examinations were performed every month for the first 6 months and every 3 months thereafter. Each examination included complete ophthalmologic evaluation, A- and B-scan ultrasonography, and IOP measurement. Sites of injections underwent regular slit-lamp examination and photographic follow-up. Liver enzyme analysis and ultrasonography were performed at 6-month intervals.
Statistical analyses (ANOVA, Kaplan-Meier, χ 2 test, independent-samples t test) were performed by SAS v.8.2 statistical package (SAS Institute, Cary, North Carolina, USA). A value of P < .05 was considered statistically significant.
Ninety-six patients (mean age, 56 ± 15 years; range, 30–86 years) were included in this retrospective study. Forty-six patients were male (46%) and 50 female (54%). The right eye was affected in 58 patients (60%) and the left eye in 38 (40%). Seventy-five melanomas were purely choroidal in location (78%), whereas 21 eyes were affected by ciliary body tumors (22%). Mean tumor largest basal diameter was 13.3 ± 3.4 mm (range, 7.6–18 mm) and mean tumor thickness was 6.6 ± 1.9 mm (range, 4–10 mm). Mean exudative retinal detachment largest basal diameter was 15.1 ± 3.5 mm (range, 10–21 mm). The rupture of Bruch membrane was present in 41 patients (43%). Mean dose rate and mean dose at tumor apex were not statistically different in the 3 groups ( P > .05) ( Table 1 ). A mean of 3.7 (range, 3–5) intravitreal injections were performed in the intravitreal bevacizumab group vs 1.9 (range, 1–3) in the intravitreal triamcinolone acetonide group. No local or systemic complications were observed during surgical procedure. Mean follow-up was 37 ± 7 months (range, 24–55 months). Marked exudative retinal detachment regression was documented in 22 of 32 intravitreal triamcinolone acetonide–treated patients (69%) vs 11 of 32 intravitreal bevacizumab–treated patients (34%) and 9 of 32 control patients (28%) (observation group) ( Table 2 ). Of these, regression was complete (absence of any ultrasonographic evidence of exudative retinal detachment) in 18 intravitreal triamcinolone acetonide–treated patients (82%) vs 10 intravitreal bevacizumab–treated patients (91%) and 7 control patients (78%) (observation group). Exudative retinal detachment marked regression was observed 4.5 ± 2.4, 7.9 ± 2.7, and 11.2 ± 3.5 months after brachytherapy, respectively. Statistical analysis showed significant difference between exudative retinal detachment marked regression in the intravitreal triamcinolone acetonide group vs intravitreal bevacizumab and observation groups ( P = .0007 and P = .0001, respectively) ( Figures 1 and 2 ). No statistical significance was found between the intravitreal bevacizumab group vs observation group ( P = .4501) ( Figures 1 and 2 ). No statistically significant difference was observed between tumors showing exudative retinal detachment regression or persistence vs patient age, sex, tumor location, exudative retinal detachment largest basal diameter, and presence of Bruch membrane rupture ( Table 3 ). Steroid-induced cataract was observed in 4 of 32 intravitreal triamcinolone acetonide–treated patients (12%). Neither steroid-induced increased intraocular pressure nor other short- or long-term side effects were documented. Seven of 96 patients (7.3%) developed neovascular glaucoma (18 ± 3.9 months after brachytherapy). Of these, 3 were in the observation group (43%), 3 were in the intravitreal bevacizumab group (43%), and 1 was in the intravitreal triamcinolone acetonide group (14%) ( P > .05). Eleven of 96 patients developed metastatic disease (11.5%) within an interval of 27 ± 5.7 months after brachytherapy. Of these, 4 were in the observation group (36%), 3 were in the intravitreal bevacizumab group (27%), and 4 were in the intravitreal triamcinolone acetonide group (36%) ( P > .05). Neither local recurrence nor extrascleral/orbital extension was documented during follow-up.
|Intravitreal Triamcinolone Acetonide Group||Intravitreal Bevacizumab Group||Observation Group||All Groups|
|Mean injections/group||1.9 (range 1–3)||3.7 (range 3–5)||NA||2.9 a (range 1–5)|
|Exudative retinal detachment: marked regression, n (%)||22 (69%)||11 (34%)||9 (28%)||42 (44%)|
|Exudative retinal detachment regression time (mo)||4.5 ± 2.4||7.9 ± 2.7||11.2 ± 3.5||6.8 ± 3.9|
|Follow-up (mo)||39 ± 8 (range 24–55)||35 ± 6 (range 24–53)||36 ± 6 (range 24–53)||37 ± 7 (range 24–55)|