To compare the treatment outcomes of a combination of pneumatic displacement and intravitreal anti–vascular endothelial growth factor, and anti–vascular endothelial growth factor monotherapy for submacular hemorrhage resulting from exudative age-related macular degeneration.
Retrospective, comparative, interventional case series.
Forty eyes treated with a combination therapy and 42 eyes treated with monotherapy for submacular hemorrhage resulting from exudative age-related macular degeneration with no significant difference in baseline central foveal thickness were compared. Central foveal thickness and best-corrected visual acuity (BCVA) at baseline, 1, 3, and 6 months after initial treatment were measured and compared between the 2 groups after adjustment of baseline central foveal thickness.
Central foveal thickness ( P < .0001) and BCVA (combination, P < .0001; monotherapy, P = .022) were improved after both treatments. Combination therapy showed more rapid improvement of central foveal thickness ( P = .009) and BCVA ( P = .007) within 1 month than monotherapy, but there was no difference at 6 months ( P = .385 and P = .303, respectively). In eyes with subretinal hemorrhage thicker than 450 μm, visual outcome at 6 months was better in the combination therapy group than in the monotherapy group ( P = .021), whereas BCVA showed no significant difference between groups in eyes with subretinal hemorrhage less than 450 μm ( P = .930).
Both treatments are useful options for submacular hemorrhage resulting from exudative age-related macular degeneration. Combination therapy may yield a better treatment outcome than monotherapy in eyes with thick subretinal hemorrhage. Nevertheless, the potential for adverse events resulting from pneumatic displacement should be considered.
Submacular hemorrhage associated with exudative age-related macular degeneration (AMD) is a complication with potentially devastating effects on visual acuity. Submacular hemorrhage frequently results in poor visual outcome because of the retinal toxicity of iron released from hemoglobin, its role as a physical barrier on diffusion of nutrients and metabolites, as well as the shearing of the outer segment photoreceptors from fibrin. Subretinal fibrosis and disciform scar formation result in further visual loss. Multiple treatments have been suggested, including pneumatic displacement with or without tissue plasminogen activator (tPA), pars plana vitrectomy with or without subretinal tPA, photodynamic therapy, and intravitreal anti–vascular endothelial growth factor (VEGF) injection. However, there remains no consensus on the optimal management of patients with submacular hemorrhage.
Anti-VEGF agents can preserve or improve vision significantly and are used widely for treatment of exudative AMD. Although patients with significant submacular hemorrhage were excluded from recent major randomized clinical trials, there is a good rationale for the efficacy of anti-VEGF therapy in such cases, including the effects on underlying neovascular membrane, the clearing effect of intravitreal and submacular hemorrhage, and reducing the risk of recurrent hemorrhage and disciform scarring. Several studies have suggested that anti-VEGF agents in combination with pneumatic displacement can improve the visual acuity in eyes with submacular hemorrhage secondary to exudative AMD. In addition, recent studies have indicated improvement of visual acuity in submacular hemorrhage treated with anti-VEGF alone.
Based on the mechanisms underlying these treatment options, a combination of pneumatic displacement with an anti-VEGF agent is used more frequently than anti-VEGF monotherapy for eyes with thicker submacular hemorrhage involving the fovea. Because of this difference in baseline characteristics, it is difficult to compare the outcomes of these treatment groups, and little information is available regarding whether the combination of displacement techniques with anti-VEGF provides any advantage over anti-VEGF alone. One study suggested that a combination of tPA and pneumatic displacement plus an anti-VEGF agent significantly improved visual acuity, whereas anti-VEGF monotherapy only stabilized the visual acuity. In contrast, another study indicated no significant differences in visual acuity or central foveal thickness between ranibizumab monotherapy and combination therapy at 12 months after initial treatment. In the present study, we analyzed a larger number of patients with thicker submacular hemorrhage compared with previous studies and compared the treatment outcome of combined pneumatic displacement plus anti-VEGF with anti-VEGF monotherapy for submacular hemorrhage associated with exudative AMD.
This retrospective, comparative, interventional case series study was performed at a single center (Severance Hospital, Seoul, South Korea) from January 2009 through March 2014. The research adhered to the tenets of the Declaration of Helsinki, and approval for the study was granted by the Institutional Review Board/Ethics Committee of Yonsei University (IRB no. 4-2014-0474). All patients provided their written informed consent.
A computerized search (clinical data repository system) was performed to identify patients diagnosed with both exudative AMD and submacular hemorrhage. We reviewed the medical data of these patients treated with anti-VEGF injection, either in combination with pneumatic displacement or as monotherapy. We analyzed patients who (1) were newly diagnosed with exudative AMD; (2) exhibited submacular hemorrhage involving the fovea of at least 1 disc diameter; (3) were treated with either a combination of pneumatic displacement and intravitreal anti-VEGF (combination therapy group) or anti-VEGF monotherapy (monotherapy group) at initial presentation, followed by 2 monthly intravitreal anti-VEGF injections; and (4) had a minimum follow-up period of 6 months. We excluded patients (1) who were younger than 50 years; (2) with submacular hemorrhages resulting from causes other than exudative AMD; (3) whose eyes were treated with tPA or photodynamic therapy during the 6-month follow-up period; (4) whose eyes had undergone vitrectomy surgery before initial presentation; (5) with evidence of end-stage AMD with severe scarring or atrophy at initial presentation; (6) whose duration of symptoms as a result of submacular hemorrhage was longer than 1 month or who had eyes with an old, yellowish discolored submacular hemorrhage; and (7) whose eyes had significant media opacity, including severe vitreous hemorrhage, at initial presentation.
During the study period, 189 eyes initially had submacular hemorrhage secondary to exudative AMD at presentation. After excluding eyes that had undergone vitrectomy at or before the initial presentation (20 eyes), photodynamic therapy during the follow-up period (7 eyes), use of tPA (4 eyes), less than 6 months of follow-up (15 eyes), severe scarring or atrophy at presentation (14 eyes), and other reasons (6 eyes), 76 eyes were treated with a combination of pneumatic displacement and anti-VEGF injection and 47 eyes were treated with anti-VEGF monotherapy. Because baseline central foveal thickness was different between the 2 treatment groups (886.2 ± 377.9 μm and 540.0 ± 219.3 μm, respectively; P < .0001), only eyes showing an overlapping range of central foveal thickness, between 380 and 850 μm, were included in the analysis. Finally, 40 eyes of 40 patients treated with a combination of pneumatic displacement and anti-VEGF (combination therapy group) and 42 eyes of 42 patients treated with anti-VEGF (monotherapy group) were analyzed.
Diagnoses of exudative AMD and submacular hemorrhage were based on the results of fundus examination, fluorescein angiography, indocyanine green angiography (ICGA; HRA2 [Heidelberg Engineering GmbH, Dossenheim, Germany]), and spectral-domain (SD) optical coherence tomography (OCT; Spectralis HRA+OCT [Heidelberg Engineering GmbH]) at either initial presentation or after resolution of hemorrhage. Patients were classified as having 1 of the 2 subtypes of exudative AMD: PCV or typical exudative AMD. PCV was diagnosed based on ICGA findings, on the presence of a branched vascular network, and on evidence of terminal polypoidal lesions, subpigment epithelial layer orange-red protrusions corresponding to the polypoidal lesions revealed by ICGA, or both. All other patients were considered to have typical neovascular AMD.
Baseline Examination and Treatment
All patients underwent comprehensive ophthalmologic examinations, including best-corrected visual acuity (BCVA) using the Snellen visual acuity chart, slit-lamp biomicroscopy, indirect ophthalmoscopy, color fundus photography, fluorescein angiography, ICGA, and OCT at the initial visit. Follow-up visits were arranged monthly for BCVA, indirect ophthalmoscopy, and OCT. Fluorescein angiography and ICGA additionally were performed after the 3 monthly injections.
At initial presentation, all patients were given intravitreal injection of anti-VEGF agent, either 0.5 mg ranibizumab (Lucentis; Novartis, Basel, Switzerland) or 1.25 mg bevacizumab (Avastin; Genentech, Inc, South San Francisco, California, USA), under sterile conditions. In the combination group, intravitreal gas injection (either SF 6 or C 3 F 8 ; mean volume, 0.3 mL) also was performed, and patients were instructed to remain in the prone position most of the time for at least 3 days. Thereafter, all patients received 2 additional anti-VEGF injections each month. After 3 monthly loading injections, patients received additional anti-VEGF injection over the following 3-month period as necessary, if persistent or recurrent subretinal or intraretinal fluid was evident, new macular hemorrhage developed, or the extent of pigment epithelial detachment increased.
Main Outcome Measures
The main outcome measures included BCVA and central foveal thickness at baseline and at 1, 3, and 6 months after the initial diagnosis. Thicknesses of the retina and submacular hemorrhage (either subretinal or sub-RPE hemorrhage) were measured at baseline and at 1 month after the initial treatment. For statistical analysis, Snellen BCVA was converted to logarithm of the minimal angle of resolution units.
Central foveal thickness, retinal thickness, and thickness of submacular hemorrhage were measured manually on the horizontal and vertical line scans intersecting the center of the fovea using the built-in caliper function of the SD OCT device. Central foveal thickness was defined as the distance between the internal limiting membrane and the Bruch membrane at the foveal center. We classified submacular hemorrhage into subretinal hemorrhage and subretinal pigment epithelium (RPE) hemorrhage according to the location of the hemorrhage. The thickness of subretinal hemorrhage was defined as the distance between the inner segment–outer segment line to the outer border of RPE, and the thickness of sub-RPE hemorrhage was defined as the distance from the inner border of RPE to the Bruch membrane, or to the outer border of fibrotic tissue at the fovea if definite sub-RPE fibrotic tissue was observed. The presumed RPE line obtained from a clearly visible RPE line was used when the RPE or Bruch membrane was not visible because of overlying thick hemorrhage. Retinal thickness was defined as the distance between the internal limiting membrane and the inner segment–outer segment line at the center of the fovea. The extent of hemorrhage was measured at baseline in square millimeters on a reference scan image obtained with SD OCT using the built-in caliper function. Thickness exceeding 1500 μm was considered as 1500 μm, and extent of hemorrhage exceeding 80 mm 2 on OCT images was regarded as 80 mm 2 , because of the possibility of inaccurate measurement. All measurements and diagnoses were conducted by 2 retinal specialists (J.Y.S. and J.L.), and the average values were used for evaluation. The main outcome measures were analyzed as outlined below.
Changes in Central Foveal Thickness and Best-Corrected Visual Acuity Over the Follow-up Period
Central foveal thickness and BCVA measured at 1, 3, and 6 months after the initial treatment were compared with the respective baseline measurements in each treatment group. For comparison of treatment outcome between the groups, a linear mixed model was used with adjustment for baseline central foveal thickness. Post hoc analysis was performed to compare the treatment outcome at each time point and to compare the changes in central foveal thickness and BCVA from baseline.
Changes in Retinal Thickness and Submacular Hemorrhage Thickness
We measured the thickness of each structure at the foveal center separately—retina, subretinal hemorrhage, and sub-RPE hemorrhage—at baseline and 1 month after the initial treatment. Comparison of these measurements between baseline and 1 month after the initial treatment was performed using a paired t test, and a comparison of changes in these measurements between combination therapy and monotherapy groups was performed using an independent t test.
Comparison of Visual Outcome According to Central Foveal Thickness and Subretinal Hemorrhage Thickness
In the present study, the association between visual outcome at 6 months and baseline central foveal thickness and the association between visual outcome at 6 months and baseline subretinal hemorrhage thickness (locally weighted scatterplot smoothed curve; Supplemental Figure , available at AJO.com ) showed that the differences in visual outcome between treatments tended to be greater in eyes with baseline central foveal thickness and baseline subretinal hemorrhage thickness values exceeding certain cutoff points. Therefore, cutoff points of baseline central foveal thickness and baseline subretinal hemorrhage thickness were chosen such that the visual outcome would be considered to be different between treatments. The data sets were searched for the point that best separated the 2 groups, which then was selected as the cutoff point the value that maximized the standard t statistics. If statistically significant, this was used as the cutoff point. Eyes in each treatment group were divided into 2 subgroups according to the cutoff point. The BCVA at baseline and at 1 and 6 months after the initial treatment were compared between these subgroups.
Other Statistical Analyses
Patient characteristics, including age, sex, duration of symptoms, antithrombotic or anticoagulant use, lens status (phakic or pseudophakic), anti-VEGF agents, and the number of anti-VEGF injections during the follow-up period, were retrieved from medical charts. For comparison of variables between the combination and monotherapy groups, an independent t test and Fisher exact test were used. Recurrence of submacular hemorrhage or adverse events after treatment, including vitreous hemorrhage, retinal detachment, or endophthalmitis, also was retrieved from medical charts. Statistical analysis was performed using SAS software version 9.2 (SAS Institute, Inc, Cary, North Carolina, USA). In all analyses, P < .05 was taken to indicate statistical significance.
There were no significant differences between the combination therapy group and monotherapy group in terms of demographics or baseline characteristics ( Table 1 ). The mean duration of symptoms was 11.4 ± 10.4 days in the combination therapy group and 13.8 ± 11.5 days in the monotherapy group ( P = .342). Baseline central foveal thickness was 590.9 ± 157.9 μm in the combination therapy group and 560.1 ± 152.7 μm in the monotherapy group. There were no differences in baseline BCVA or central foveal thickness between the 2 groups ( P = .670 and P = .386, respectively). Diagnosis (either typical AMD or PCV) also was not different between the 2 treatment groups ( P = .376).
|Characteristic||Pneumatic Displacement With anti-VEGF||Anti-VEGF Monotherapy||P Value|
|Age (y)||72.0 ± 8.3||74.6 ± 6.8||.127|
|Male sex, no. (%)||26 (65.0)||27 (64.3)||>.999|
|Duration of symptom (d)||11.4 ± 10.4||13.8 ± 11.5||.342|
|Antithrombotic/anticoagulant use, no. (%)||7 (17.5)||10 (23.8)||.589|
|Phakic lens status, no. (%)||33 (82.5)||34 (81.0)||>.999|
|No. of anti-VEGF injections||4.3 ± 1.1||4.6 ± 1.2||.345|
|Diagnosis, no. (%)|
|Typical AMD||21 (52.5)||17 (40.5)||.376|
|PCV||19 (47.5)||25 (59.5)|
|BCVA (logMAR)||1.21 ± 0.71||1.14 ± 0.63||.670|
|Central foveal thickness (μm)||590.9 ± 157.9||560.1 ± 152.7||.386|
|Extent of hemorrhage (mm 2 )||23.5 ± 21.8||23.4 ± 22.3||.997|
For initial treatment, 23 eyes in the combination therapy group and 29 eyes in the monotherapy group were treated with ranibizumab ( P = .360). In the combination therapy group, pneumatic displacement was performed with SF 6 in 20 eyes and C 3 F 8 in 20 eyes. The number of anti-VEGF injections within 6 months was not significantly different between the 2 groups ( P = .345).
Changes in Central Foveal Thickness Over the Follow-up Period
Central foveal thickness improved significantly after both combination therapy ( Figure 1 ) and monotherapy ( Figure 2 ) during the 6-month follow-up period (both P < .0001; Figure 3 ). Central foveal thickness decreased from 590.9 ± 157.9 μm at baseline to 279.2 ± 196.2 μm at 6 months in the combination therapy group and from 560.1 ± 152.7 μm at baseline to 311.5 ± 190.2 μm at 6 months in the monotherapy group. To minimize the effect of differences in baseline central foveal thickness, comparisons between the 2 groups were performed after adjustment for the initial central foveal thickness, although baseline central foveal thickness was not significantly different between the 2 groups. From baseline to 1 month after initial treatment, decrease in central foveal thickness was significantly greater in the combination therapy group than in the monotherapy group ( P = .009), but the differences were not significant between baseline and 3 months and between baseline and 6 months after initial presentation ( P = .127 and P = .385, respectively).
Changes in Visual Acuity Over the Follow-up Period
Changes in BCVA over the follow-up period are shown in Figure 3 . Logarithm of the minimal angle of resolution BCVA improved from 1.21 ± 0.71 to 0.76 ± 0.59 at 6 months in the combination therapy group and from 1.14 ± 0.63 to 0.94 ± 0.61 at 6 months in the monotherapy group. In the combination therapy group, BCVA improved significantly from baseline to 1 month ( P < .0001), to 3 months ( P = .0003), and to 6 months ( P < .0001). Meanwhile, in the monotherapy group, BCVA showed no significant difference from baseline at 1 month ( P = .905) or 3 months ( P = .079), but showed significant improvement from baseline at 6 months after initial treatment ( P = .022). Comparisons between the 2 groups were performed after adjustment for initial central foveal thickness. BCVA was not significantly different between the 2 groups at baseline ( P > .999), 3 months ( P = .579), or 6 months ( P = .474) after initial treatment, but the combination therapy group showed better visual acuity at 1 month after initial treatment compared with the monotherapy group ( P = .017). The changes in BCVA from baseline to 1 month were significantly greater in the combination therapy group than in the monotherapy group ( P = .007), whereas there was no significant difference between the 2 groups from baseline to 3 months ( P = .361) or from baseline to 6 months after the initial treatment ( P = .303).
Changes in Retinal Thickness and Submacular Hemorrhage Thickness
The thicknesses of submacular hemorrhage and the retina were decreased significantly 1 month after initial treatment in both the combination therapy group (subretinal hemorrhage, P < .0001; sub-RPE hemorrhage, P = .018; retinal thickness, P = .001) and monotherapy group ( P < .0001, respectively). In terms of subretinal hemorrhage, changes in subretinal hemorrhage thickness were greater in the combination therapy group than in the monotherapy group ( P = .002). Sub-RPE hemorrhage did not show significant differences at baseline or 1 month after initial treatment ( P = .508 and P = .987, respectively), and changes in thickness of sub-RPE hemorrhage also were not significantly different between the 2 groups ( P = .554). Retinal thickness showed no differences at baseline ( P = .238) or 1 month after initial treatment ( P = .319), and changes in retinal thickness were not significantly different between the 2 groups ( P = .481; Table 2 ).