Anti-Vascular Endothelial Growth Factor Therapy for Stages 3 and 4 Proliferative Sickle Cell Retinopathy Results in Improved Anatomic and Visual Outcomes

Abstract

Purpose:

To investigate the potential utility of intravitreal anti-vascular endothelial growth factor (anti-VEGF) therapy for stages 3 and 4 proliferative sickle cell retinopathy (PSR).

Methods:

Retrospective review of anatomic and visual acuity outcomes after intravitreal anti-VEGF therapy for stages 3 and 4 PSR eyes.

Results:

There were 45 PSR eyes (17 stage 3 and 28 stage 4) treated with anti-VEGF agents (bevacizumab (37 eyes), aflibercept (6 eyes), ranibizumab (3 eyes), one eye received 2 both bevacizumab and aflibercept). Follow-up ranged from 1 to 120 months (mean= 21 months). Within 1 month after anti-VEGF treatment, VA improved ≥ 2 lines in 17/45 eyes, remained stable in 27/45 eyes and worsened in 1/45 eyes. Median visual acuity remained 20/20 for stage 3 and improved from 20/200 to 20/30 for stage 4 PSR eyes. More stage 4 PSR eyes had VA improvement post-treatment than stage 3 PSR eyes (P=0.003). There were no cases of endophthalmitis. Visual acuity outcomes were similar for anti-VEGF with and without laser treatment. Control of PSR was achieved in 42/45 (93%) of eyes without the need for pars plana vitrectomy.

Conclusion:

Within one month of treatment, administration of anti-VEGF therapy of PSR eyes resulted in seafan regression, clearing of vitreous hemorrhage, and stable or improved vision in most eyes.

Summary statement:

Anti-VEGF therapy, with or without laser photocoagulation, for proliferative sickle cell retinopathy (PSR) can result in anatomic control with visual acuity stability for stage 3 and marked improvement in stage 4 eyes.

Proliferative sickle cell retinopathy (PSR) is defined by the presence of retinal neovascularization and includes Goldberg PSR stages 3 to 5.^1–3 Untreated, stage 3 PSR, characterized by the development of retinal neovascularization in the classic seafan configuration, can progress to vitreous hemorrhage (stage 4 PSR) and retinal detachment (stage 5 PSR). A prospective clinical trial established laser photocoagulation was effective for prevention of visual loss because it lowered the incidence of vitreous hemorrhage in stage 3 PSR eyes compared to observation.^4–5 Several case series have shown efficacy of pars plana vitrectomy (PPV) for treatment of stage 4 PSR.^6–10 More recently, case reports suggest utility of anti-vascular endothelial growth factor (anti-VEGF) intravitreal injections for treatment of stages 3 and 4 PSR eyes.^11–17 To date, there have been neither large case series nor a phase 3 randomized clinical trial comparing the efficacy of anti-VEGF to laser photocoagulation for stage 3 PSR and no study comparing anti-VEGF to observation for stage 4 PSR. The Anti-VEGF versus Photocoagulation for Sickle Cell Retinopathy Study (ALPS) is a prospective, randomized, controlled clinical trial that will investigate the utility of anti-VEGF therapy for stages 3 and 4 PSR. We surveyed retina specialists of the ALPS Study Group to determine their experience in usage of anti-VEGF therapy for stages 3 and 4 PSR to help guide the design of the prospective clinical trial. Six of the ALPS Study Group members reported usage of anti-VEGFs for PSR. This report summarizes their treatment approaches and outcomes.

Methods:

The ALPS Study Group of investigators received an electronic survey to retrospectively collect visual and anatomic outcomes data after anti-VEGF treatment of eyes with stages 3 and 4 PSR. Local IRB approvals were obtained for this retrospective review of anti-VEGF therapy for PSR. De-identified data were collected electronically. Data collected included baseline demographics (age, sickle cell subtype, systemic sickle cell treatment(s)), ocular history (prior treatments and surgeries), stage of PSR, anti-VEGF treatment parameters (drug name, frequency), visual acuity pre- and post-anti-VEGF therapy, anatomic outcomes (status of seafan regression and fibrosis, vitreous hemorrhage (VH), development of new seafans or new VH) and complications (endophthalmitis, retinal tears, tractional (TRD) or rhegmatogenous detachment (RRD)). Only previously non-published cases were included in this series. Visual and anatomic outcomes were noted after anti-VEGF treatment for each stage of PSR. The data were collated (JIL) and analyzed.

Results:

There were a total of 62 eyes of 59 patients, who had received an intravitreal injection of an anti-VEGF drug for either stage 3 or stage 4 PSR in the United States (51 eyes in 47 patients) or Nigeria (11 eyes in 12 patients). Of these, 17 eyes were excluded because the patient had sickle cell traits (8), concomitant diabetic retinopathy (4), choroidal neovascularization (2), branch vein occlusion with CME (1) or no follow-up (2) visit. This left 45 eyes of 43 patients for inclusion; 29 (64%) eyes were hemoglobin (Hgb) SC, 9 (20%) were HgbSS, 2 (4%) were Hgb Sthal and 5 (11%) were unspecified (all 5 patients were actively on sickle cell medications or exchange transfusions). Of the 45 eyes, 17 were stage 3 PSR (11 were Hgb SC, 5 were Hgb SS, 1 was of unknown Hgb subtype) and 28 were stage 4 PSR (18 were Hgb SC, 4 were Hgb SS, 2 were SThal and 4 were of an unknown Hgb subtype). The demographics and anti-VEGF agents used are given in Table 1. Eighteen of the 45 (40%) eyes had received prior laser photocoagulation, and 2 (4%) eyes had undergone PPV. The intravitreal agents used included bevacizumab (37 yes), aflibercept (6 eyes) and ranibizumab (3 eyes); one eye received both aflibercept and bevacizumab at various times.

Table 1.

Baseline Demographics

PSR Stage	N (eyes)	N (Pts)	Mean Age (years)	Hgb SC (%)	Hgb SS (%)	Hgb Sthal (%)	Hgb Unknown (%)
All Stages	45	43	43	29 (64)	9 (20)	2 (4)	5 (11)
Stage 3	17	17	43	11 (65)	5 (29)	0	1 (6)
Stage 4	28	28	45	18 (64)	4 (14)	2 (7)	4 (14)

PSR Stage	N (eyes)	Bevacizumab N (%)	Aflibercept N (%)	Ranibizumab N (%)
All Stages	45	37* (82)	6* (13)	3 (7)
Stage 3	17	12 (71)	3 (18)	2 (12)
Stage 4	28	25 *(86)	3* (11)	1 (4)

• One eye received both bevacizumab and aflibercept during the course of treatment

Follow-up ranged from 1 to 120 months and mean follow-up was 21 months (median follow-up was 18 months). There was at least one year of follow-up for 28/45 (62%) eyes (15 stage 4 PSR and 13 stage 3 PSR eyes). Follow-up was less than one year for 14/45 (31%) eyes; limited to 1 month for 2 eyes, 3 months for 7 eyes and 6 months for 5 eyes.

Anti-VEGF treatment regimens varied amongst the treating physicians. Treatment regimens for stage 3 PSR included anti-VEGF alone (8) and anti-VEGF followed by laser photocoagulation (9, of which 7 had laser within 6 weeks of the anti-VEGF administration (early laser) and 2 had laser more than 6 weeks after the anti-VEGF administration (delayed laser)). Treatment regimens for stage 4 PSR included anti-VEGF alone (14), anti-VEGF with laser within 6 weeks (9), anti-VEGF with delayed laser (2) and anti-VEGF with delayed PPV (3).

Prior to anti-VEGF therapy, laser was present in 15/28 (54%) stage 4 PSR eyes and 3/17 (18%) stage 3 PSR eyes (Chi square = 5.7, P=0.02). Repeat anti-VEGF therapy was performed in 21 of 45 (47%) eyes. Repeat anti-VEGF tended to be performed more often for stage 4 PSR (16/28 (57%)) than stage 3 PSR eyes (5/17, (29%)) (Chi square=3.3, P=0.07). Repeat anti-VEGF tended to be performed more often for HgbSC (14/22, 64%) than for other Hgb subtypes although the difference was not significant (8/23, 35%) (Chi Square=3.7, p=0.05).

Baseline visual acuity ranged from 20/20 to LP (median 20/60). Final visual acuity ranged from 20/20 to HM (median 20/25). Visual acuity improved two or more lines in 17/45 (38%) eyes, remained stable in 27/45 (60%) and decreased in 1/45 (2%) eyes at one month. Visual acuity improved two or more lines in 18/45 (40%) eyes, remained stable in 25/45 (56%) eyes and decreased in 2/45 (4%) eyes at the final visit. The visual acuity response by treatment regimen is given in Table 2. PSR was controlled without the need for PPV in 42/45 eyes (93%). Half of stage 3 PSR and half of the stage 4 PSR eyes were managed with anti-VEGF treatment alone. The remaining eyes received anti-VEGF followed by laser photocoagulation in the treatment course. The treatment regimens and rationale for adding laser or PPV varied across institutions and were not standardized.

Table 2.

Visual acuity response by anti-VEGF treatment group

		Visual Acuity Change from Baseline at Month 1			Visual Acuity Change from Baseline at Final Visit
Treatment Group	N (eyes)	≥ 2 lines Increase (%)	Stable (%)	≥ 2 lines Decrease (%)	≥ 2 lines Increase (%)	Stable (%)	≥ 2 lines Decrease (%)
Anti-VEGF (All groups)	45	17 (38)	27 (60)	1 (2)	18 (40)	25 (56)	2 (4)
Anti-VEGF alone	22	7 (32)	15 (68)	0	9 (41)	12 (55)	1 (5)
Anti-VEGF + laser	16	6 (38)	10 (63)	0	6 (38)	10 (63)	0
Anti-VEGF then laser	4	2 (50)	2 (50)	0	2 (50)	2 (50)	0
Anti-VEGF then PPV	3	2 (67)	0	1 (33)	1 (33)	1 (33)	1 (33)

Stage 3 PSR Eyes

For eyes with stage 3 PSR, anti-VEGF alone or combined anti-VEGF plus laser was used. The combination treatment group, divided into an early combination group (defined previously as laser within 6 weeks of the initial anti-VEGF treatment), and a delayed combination group (defined previously as laser 6 or more weeks after the initial anti-VEGF treatment). Baseline visual acuity ranged from 20/20 to 20/200 with a median of 20/20 and remained 20/20 at month 1 and at the last visit (ranging from 20/20 to hand motions (HM)). Follow-up ranged from 1 to 36 months; mean follow-up was 31 months (median was 24 months). The change in visual acuity from baseline by anti-VEGF treatment group for stage 3 PSR eyes is given in Table 3. The majority of eyes (88%) were stable at month 1 and at the final visit (82%). There was no difference in the proportion of eyes that had improved vision versus stable vision for follow-up durations less than 6 months, 6 months to one year or one or more years (Fishers Exact, P=1) since almost all eyes remained stable compared to baseline (Table 4). Overall, safety outcomes were good. There were no cases of endophthalmitis, retinal tear formation or RRDs.

Table 3.

Change in visual acuity from baseline by anti-VEGF treatment group for PSR stage 3 eyes

		Visual Acuity Change from Baseline at Month 1			Visual Acuity Change from Baseline at Final Visit
Treatment Group	N (eyes)	≥ 2 lines Increase (%)	Stable (%)	≥ 2 lines decrease (%)	≥ 2 lines increase (%)	Stable (%)	≥ 2 lines decrease (%)
Anti-VEGF	17	2 (12)	15 (88)	0	2 (12)	14 (82)	1 (6)
Anti-VEGF alone	8	0	8	0	0	7 (88)	1 (12)
Anti-VEGF + laser early	7	2 (29)	5 (71)	0	2 (29)	5 (71)	0
Anti-VEGF + delayed laser	2	0	2	0	0	2	0

Table 4.

Change in visual acuity from baseline by anti-VEGF treatment group for PSR stage 4 eyes

		Visual Acuity Change from Baseline at Month 1			Visual Acuity Change from Baseline at Final Visit
Treatment Group	N (eyes)	≥ 2 lines Increase (%)	Stable (%)	≥ 2 lines Decrease (%)	≥ 2 lines Increase (%)	Stable (%)	≥ 2 lines Decrease (%)
Anti-VEGF	28	16 (57)	11 (39)	1 (4)	16 (57)	11 (39)	1 (4)
Anti-VEGF alone	14	8 (57)	6 (43)	0	9 (64)	5 (36)	0
Anti-VEGF + laser	9	4 (44)	5 (56)	0	4 (44)	5 (56)	0
Anti-VEGF + delayed laser	2	2 (100)	0	0	2	0	0
Anti-VEGF then PPV	3	2 (67)	0	1 (33)	1 (33)	1 (33)	1 (33)

There were 8 of 17 (47%) eyes managed with anti-VEGF alone: visual acuities for all 8 eyes were stable or improved at one month; 7 of 8 (88%) eyes had stable visual acuities and one of 8 (12%) eyes had worse visual acuity at the final follow-up. The patient who had worse final compared to baseline vision had baseline visual acuity of 20/400, which gradually declined to HM because of ischemia. For eyes in the early combination group, all 7 eyes were stable (5 eyes) or improved (2 eyes) at both 1 month and at the final visit. For eyes in the late combination group, 2 of 2 eyes were stable at both one month and at the final follow-up visit. Seafans were controlled with these treatment regimens.

For the stage 3 PSR eyes, complete regression of seafans occurred after only one anti-VEGF injection (without any laser treatment) in 8 eyes. Of these 8 eyes, 3 (38%) experienced recurrence of seafans. Recurrences occurred at month 5 for one of the three patients, month 7 for another one of the three patients and at month 18 for another one of the three patients. Recurrent seafans requiring repeat anti-VEGF therapy occurred in one of 9 (11%) eyes which had been treated with anti-VEGF followed by laser.

Overall, of the 17 total stage 3 PSR eyes receiving anti-VEGF therapy, laser was used as additional therapy for seafan activity after the initial treatment in 3 eyes (18%); of these, 2 of the 3 eyes also received additional anti-VEGF therapy. The third patient declined additional intravitreal anti-VEGF injections due to out-of-pocket drug cost.

There were no differences in visual acuity outcomes at one month or at the final visit between eyes treated with anti-VEGF alone versus anti-VEGF plus laser. At month 1, for anti-VEGF treatment alone, none of the 17 (0%) eyes improved two or more lines of vision and all 8 (100%) were stable compared to eyes in the anti-VEGF plus laser group, of which 2 of 7 (29%) improved and 5 of 7 (71%) were stable (FE, P=0.2). At the final visit, for the anti-VEGF treatment alone group, none improved and 7/7 (100%) were stable compared with the anti-VEGF plus laser group of which 2 of the 7 (29%) improved and 5 of the 7 (71%) eyes were stable at the final visit (FE, P=0.46).

PPV was performed in 1 (6%) of the original stage 3 PSR eyes six years after the first anti-VEGF injection because of a new vitreous hemorrhage. New seafans occurred in 2 of17 (12%) eyes; one patient had a new seafan at month 5 (anti-VEGF at baseline and month 5) and the other patient had a new seafan at month 18 (anti-VEGF had been given at baseline and then at months 18 and 21 followed by laser at months 2 and 20).

Stage 4 PSR Eyes

For eyes with stage 4 PSR, anti-VEGF therapy was repeated as needed until vitreous hemorrhage either cleared or until the hemorrhage cleared enough to permit laser photocoagulation of the seafans. Of the 28 eyes with vitreous hemorrhage, 14 (50%) eyes received anti-VEGF alone. Eleven of the 28 (39%) eyes were treated with anti-VEGF and laser photocoagulation, and 3 (11%) underwent anti-VEGF followed by PPV. Follow-up ranged from 1 to 36 months; mean follow-up was 15 months (median, 6 months).

Overall, safety outcomes were good. There were no cases of endophthalmitis, retinal tear formation or RRDs.

Baseline visual acuity ranged from 20/20 to light perception (LP). Median visual acuity was 20/200 at baseline and improved to 20/30 by month 1 and at the last follow-up (range 20/20 to counts fingers (CF)). Change in visual acuity from baseline by anti-VEGF treatment group for stage 4 PSR eyes is given in Table 4. Visual acuity at month 1 was improved compared to baseline for a greater proportion of stage 4 PSR eyes compared to stage 3 PSR eyes (Chi square = 9.1; P=0.003). Within 1 month after anti-VEGF therapy, VA improved two or more lines in 16 of 28 (57%) eyes, remained stable in 11 of 28 (39%) eyes and worsened in one of 28 (4%) eyes (20/40 baseline decreasing to HM). Seven of the 11 eyes with stable visual acuity had baseline visual acuity of 20/20– 20/30 and one eye each had baseline visual acuity of 20/125, 20/200, HM or LP.

Final visual acuity was improved compared to baseline for a greater proportion of stage 4 PSR eyes compared to stage 3 PSR eyes (Chi square = 9.1; P=0.003). There was no difference in the proportion of eyes that had improved vision versus stable vision for follow-up durations less than 6 months, 6 months to one year or 1 or more years (Table 5, FE, p=0.6). At the last follow-up, VA was improved two or more lines in 16 of 28 (57%) eyes, remained stable in 11 of 28 (39%) and worsened in one of 28 (4%). Of the 11 eyes with stable vision, 7 eyes had visual acuity ranging from 20/20 to 20/30 and one eye each had 20/125, 20/200, 20/400 or HM vision at baseline. For the one patient whose final visual acuity was worse compared to baseline, baseline visual acuity declined from 20/40 to 20/70. Final visual acuity was less than 20/40 in five of the 28 eyes. These five eyes included one eye with CF visual acuity at month 24 due to aphakia and posterior capsular opacification (PCO) (1), one eye with vitreous hemorrhage at month 3 and no further follow-up (1), one eye with a TRD, which underwent PPV and silicone oil with subsequent formation of a PCO and visual acuity of counts fingers at month 24 (1), one eye with a TRD, which underwent PPV and developed postoperative ERM and CME, resulting in 20/70 vision at month 24 (1) and one eye that had ERM with CME that has not yet had PPV.

Table 5.

Change in visual acuity from baseline by anti-VEGF treatment group and length of follow-up for Stage 3 and Stage 4 PSR Eyes

		Change in visual acuity from baseline
		Stage 3 PSR Eyes			Stage 4 PSR Eyes
Length of Follow-up	N (eyes)	≥ 2 lines Increase (%)	Stable (%)	≥ 2 lines Decrease (%)	≥ 2 lines Increase (%)	Stable (%)	≥ 2 lines Decrease (%)
< 6 months	10	0	2	0	6	2	0
> 6 to 12 months	7	0	2	0	3	2	0
> 1 year	28	2	11	0	7	6	2

For stage 4 PSR eyes with baseline VA 20/40 or worse, 15/21 (71%) improved 2 or more lines by month 1, 4 (19%) remained stable and one (5%) worsened. At the last follow-up, VA was improved 2 or more lines in 16 of21 (76%) eyes, remained stable in 4 of 21 (19%) eyes (one eye each with 20/125, 20/400, HM visual acuity at baseline) and worsened in one of 21 (5%) eye (20/40 to 70).

PPV was performed in 3 of 28 (11%) eyes after anti-VEGF treatment. Two of the 3 eyes had a TRD diagnosed upon clearing of the vitreous hemorrhage; both were successfully reattached with PPV. The third eye underwent PPV because of the patient’s complaints of floaters due to residual vitreous opacities related to vitreous hemorrhage. One of the two TRD eyes had improved from light perception pre-operatively to 20/70 postoperatively at month 24; visual acuity was limited by an ERM and CME. In the other TRD eye of another patient, visual acuity improved from hand motions (baseline counts fingers) to counts fingers at month 24; visual acuity was limited due to band keratopathy, silicone oil and PCO. For the patient underwent who underwent PPV for vitreous floaters, visual acuity improved from 20/40 preoperatively to 20/25 postoperatively by month six.

There were no differences in visual acuity outcomes at one month or at the final visit between eyes treated with anti-VEGF alone versus anti-VEGF plus laser. At month 1, for anti-VEGF treatment alone, 8 of 14 (57%) eyes improved 2 or more lines and 5 of 14 (43%) were stable compared to anti-VEGF plus laser for which 6 of11(55%) improved and 5 of 11 (45%) were stable (FE, p=1). At the final visit, for anti-VEGF treatment alone, 9 of 14 (64%) improved and 5 of 14 (38%) were stable compared to anti-VEGF plus laser, 6 of 11 (55%) improved and 5 of 11 (45%) were stable at the final visit (FE, p=0.7).

There were no cases of endophthalmitis.

Discussion:

Over the last four decades, there have been no clinical trials investigating new treatments for the management of PSR. Laser photocoagulation remains the only extensively studied and proven therapy. Case reports suggest utility of anti-VEGF therapy for stages 3 and 4 PSR^11–17. Anti-VEGF therapy has been proven to be effective in the treatment of retinal neovascularization in eyes with proliferative diabetic retinopathy and retinal vein occlusion, choroidal neovascularization (age-related, myopic and idiopathic) and macular edema related to retinal vascular diseases. In contrast, the evidence of efficacy for anti-VEGF therapy for PSR includes only retrospective case reports with limited follow-up. The information from our retrospective study provides additional evidence for the clinical utility of anti-VEGF therapy for stages 3 and 4 PSR.

In 1999, Cao et al used immunohistochemistry to show VEGF and bFGF were highly associated with seafans on cryosections of postmortem retinal specimens of patients with SCR.¹⁸ They showed both nonproliferative and proliferative SCR eyes had increased VEGF and bFGF in the retina compared with non-sickle control eyes (p<0.01 and p<0.02 respectively). However, they did not find increased VEGF in the ischemic retina nor evidence increased vascular permeability. These findings led to the conclusion that there was an autocrine model for VEGF secretion in SCR eyes as compared to a paracrine model, which is seen in diabetic retinopathy.¹⁸

In 2005, Mohan et al showed involvement of the angiopoietin/Tie-2 system in PSR eyes. The researchers found elevated plasma levels of angiopoietins (Ang-1, Ang-2), VEGF, and von Willebrand factor (vWf) in the plasma of sickle cell disease patients compared with normal controls.¹⁹ The cytokine levels, however, did not differ according to severity of SCR and only limited changes occurred after laser photocoagulation.¹⁹

In 2016, Rodrigues et al, in a postmortem study, showed increased levels of hypoxia-inducible factor 1α (HIF-1a) and VEGF in the inner retina of untreated PSR stage 3 eyes (5/5 eyes) adjacent to seafans and no expression of HIF-1α (and VEGF only lightly detected) in normal retinal and choroidal vasculature of control eyes (3/3 eyes).²⁰ The investigators also found HIF-1a and VEGF strongly expressed in retinal cells within the avascular (nonperfused) retina, anterior to the boundary between perfused and nonperfused retina, and in posterior ischemic retina in the presence or absence of neovascular seafans. Their findings showed evidence for both autocrine and paracrine mechanisms for neovascularization in PSR eyes. The findings suggest that location of angiogenic growth factor secretion should drive the rationale for laser placement around sea-fan neovascularization. It is possible that anti-VEGF therapy may be useful to quench both local VEGF production and more distant ischemia-related, paracrine-driven VEGF.

Our series represents the largest series and has the longest follow-up of patients treated with of anti-VEGF therapy for stage 3 and stage 4 PSR. Our series shows evidence of safety and some efficacy. No cases of endophthalmitis or retinal vasculitis occurred. Visual acuity remained good (median visual acuity of 20/20) for stage 3 PSR eyes and markedly improved for PSR stage 4 eyes (generally had baseline visual acuity worse than 20/40) after anti-VEGF treatment. Overall, only 2 eyes in the series lost more than 2 lines of visual acuity. One eye of the 17 stage 3 PSR eyes lost 2 or more lines of visual acuity because of progressive, severe macular ischemia over a 2-year period. One eye of the 28 stage 4 PSR eyes lost more than 2 lines of visual acuity because of progression of a TRD. Eyes which showed no improvement in visual acuity had either a media opacity (PCO), ERM or persistent vitreous hemorrhage which was not removed with PPV.

Our series did include a variety of anti-VEGF agents and treatment regimens, all of which were associated with clinical regression of seafans. Anti-VEGF therapy was effective in causing regression of seafans in eyes that had undergone prior laser treatment, in eyes with recurrent seafans after initial regression and for new seafans. Repeat anti-VEGF therapy was well tolerated by patients with both PSR stage 3 and stage 4 PSR eyes. The time to seafan recurrence varied in duration after anti-VEGF therapy.

A longer term and randomized trial is needed to determine the durability of anti-VEGF treatment alone and the rate of recurrent seafan formation after anti-VEGF therapy for stage 3 PSR eyes. Our series included various anti-VEGF regimens, including immediate and delayed laser, which limit the power of the study to determine the effect of anti-VEGF therapy alone. However, analysis of the limited dataset did not show any difference in visual acuity outcomes between anti-VEGF alone and combination therapy.

Our series, however, provides additional support for the rapid clearance of vitreous hemorrhage, typically within one month, after anti-VEGF treatment. Other case reports and series have also shown fast resolution of vitreous hemorrhage with anti-VEGF treatment in stage 4 PSR eyes.^11–17 The use of anti-VEGF for stage 4 PSR could result in significant socioeconomic benefits for the patient, who could be spared PPV because of an intravitreal injection of anti-VEGF. Further work is needed to investigate the rates of seafan regression in eyes with stage 4 PSR. This current series cannot inform us of the minimum number of anti-VEGF treatments needed for stage 4 PSR eyes.

The potential benefit of using an anti-VEGF for treatment of PSR eyes includes the widespread global suppression of VEGF within the posterior segment and the potential to suppress new seafan formation as compared to focal suppression of VEGF with laser treatment. Based on the postmortem histopathological study of PSR eyes that showed increased VEGF within a 2mm area surrounding seafans, more generous applications of laser may be beneficial and aid in regression of seafans and prevention of new seafan formation.²⁰ Additionally, anti-VEGF therapy could result in less damage to the peripheral visual field compared to laser photocoagulation. The ALPS study will be able to compare anti-VEGF to laser photocoagulation for PSR stage 3 and anti-VEGF compared to observation for stage 4 PSR eyes. Such a study for stage 4 PSR would be similar to the Diabetic Retinopathy Network Protocol AB, which showed benefit for anti-VEGF for diabetic vitreous hemorrhage.²¹ The ALPS will help determine whether anti-VEGF treatment is similarly effective, less effective or more effective than laser treatment.

Our study is limited by its retrospective design with the inherent absence of a standardized anti-VEGF treatment protocol for stage 3 and stage 4 PSR eyes, lack of a control group, data selection and response bias due to the survey nature of the data, exclusion of sickle cell trait patients, and variable follow-up. Our study is not a randomized study and has neither the power to determine the optimal number of treatments nor the ability to assess the utility of combination anti-VEGF and laser photocoagulation. In addition, a comparison of the efficacy of the various anti-VEGF agents is not possible because 80% were treated with bevacizumab and no randomization was used in the treatment decision. Since we did not randomize eyes to laser or anti-VEGF, there may be biases in the treatment selected and thus we cannot state which is a better treatment or which anti-VEGF drug is the better drug. In addition, due to the very good baseline visual acuity of the stage 3 PSR eyes, the efficacy of anti-VEGF in stage 3 PSR may be understated for this group of eyes. A randomized prospective study is needed.

In summary, our series of cases show that anti-VEGF intravitreal injections can result in control of seafan neovascularization in PSR eyes with maintenance of good visual acuity, clearance of vitreous hemorrhage with marked improvements in visual acuity and evidence of safety in stage 3 and stage 4 PSR eyes. Additional research is needed to determine the relative efficacy of anti-VEGF alone compared to laser for stage 3 PSR eyes and for anti-VEGF alone compared to observation for stage 4 PSR eyes.