- Email: [email protected]
Long-Term Outcomes of Total Exudative Retinal Detachments in Stage 3B Coats Disease
Long-Term Outcomes of Total Exudative Retinal Detachments in Stage 3B Coats Disease Albert S. Li, MD,1 Antonio Capone, Jr., MD,2 Michael T. Trese, MD,2 Jonathan E. Sears, MD,3 Andres Kychenthal, MD,4 Irina De la Huerta, MD, PhD,2 Philip J. Ferrone, MD5 Purpose: To evaluate the long-term outcomes of treatment of total exudative retinal detachments (ERDs) secondary to Coats disease (stage 3B) and the role of vitrectomy. Design: Retrospective, observational case series. Participants: A total of 16 eyes in 16 patients undergoing treatment for total ERDs secondary to Coats disease with at least 5 years of follow-up. Methods: We reviewed the records of patients with stage 3B Coats disease. The interventions, including the timing of vitrectomy if used, and clinical course were recorded. Main Outcome Measures: The primary outcome measures were visual acuity at the most recent appointment, whether there was progression to neovascular glaucoma (NVG) or phthisis bulbi, and need for enucleation. Results: All patients received ablative treatment (photocoagulation or cryotherapy), with 8 having scleral buckling (SB) and 6 having external drainage of subretinal fluid (XD). Of the 12 patients who had pars plana vitrectomy (PPV), 8 had early PPV (EV) in the first year after presenting, and 4 of 8 in the expectant management group had late PPV (late vitrectomy) at a mean of 4.3 years post-presentation for treatment of significant traction retinal detachment (TRD). The other 4 patients of 8 in the expectant management group did not require vitrectomy. Mean follow-up overall was 9 1/2 years. At the date of last follow-up, 50% had no light perception or light perception vision, which was consistent across the subgroups that underwent EV (4/8), late vitrectomy (2/4), or no PPV (2/4). A total of 4 of 16 patients had progression to NVG or phthisis, 1 of whom required enucleation. Conclusions: In this retrospective series of patients with Stage 3B Coats disease, ablative therapy with a combination of PPV, XD, or SB was effective in preventing progression to NVG or phthisis in the majority of patients, thus preserving the globe. Half of the patients (4/8) in this series who did not undergo PPV in the early vitrectomy group developed late-onset TRD, suggesting a possible role for early prophylactic vitrectomy with possible SB and XD; however, this is balanced by the other half (4/8) in the expectant management group who did not require any vitrectomy. Ophthalmology 2017;-:1e7 ª 2017 by the American Academy of Ophthalmology
Coats disease is a nonhereditary disorder of the retinal vasculature, characterized by leaky retinal telangiectasia and associated subretinal or intraretinal exudation.1 It usually is present unilaterally and predominantly affects young men. There is a wide spectrum of presentation for Coats disease; it can be found as an asymptomatic retinal telangiectasia noted on examination or can present with exuberant exudation causing retinal detachment, progressing to secondary neovascular glaucoma (NVG) and phthisis. Shields et al2 proposed the definition of Coats disease as an “idiopathic retinal telangiectasia with intraretinal and/or subretinal exudation without appreciable retinal or vitreal traction.” Drawing on their experience with 150 such cases, they also presented a staging system that classifies Coats disease on the basis of the severity of these findings.2 Management of Coats disease varies according to the stage of disease. In cases of retinal telangiectasia only (stage 1) or telangiectasias with exudation (stage 2) that is ª 2017 by the American Academy of Ophthalmology Published by Elsevier Inc.
not vision threatening, close observation can be used. Laser photocoagulation and cryotherapy constitute the mainstay treatment for Coats disease once exudation becomes significant. Both are options in cases of exudative subtotal retinal detachment (stage 3A) and even exudative total retinal detachment (stage 3B).3 Nonetheless, treatment of stage 3A and 3B Coats disease is challenging because the exuberant exudation characteristic of this stage is not as responsive to ablative therapy and often requires multiple sessions to eliminate all active and recurrent Coats lesions. Stage 3 Coats disease is also on the cusp of progressing onto intractable disease, with stage 4 characterized by NVG and stage 5 marked by a blind eye with cataract and phthisis bulbi. For stage 3B, the goals of treatment are maintaining any residual vision and globe preservation without pain. Advanced Coats disease with prolific exudation can present with leukocoria and be difficult to distinguish from retinoblastoma on examination. In such cases in which https://doi.org/10.1016/j.ophtha.2017.12.010 ISSN 0161-6420/17
1
Ophthalmology Volume -, Number -, Month 2017 retinoblastoma cannot be definitively excluded, ophthalmologists have opted to enucleate an eye with poor prognosis so as not to delay diagnosis and definitive treatment of a potentially lethal condition, inevitably also enucleating some eyes with advanced Coats disease. However, in 1988, Silodor et al4 showed that surgical intervention in eyes with confirmed stage 3B Coats disease was effective in salvaging the globe for cosmesis and staving off progression to painful NVG. In their series of 13 patients with total bullous exudative retinal detachments (ERDs) from advanced unilateral Coats disease, 7 eyes were treated with intraocular infusion, external drainage of subretinal fluid (XD) and cryotherapy; the mean age at diagnosis of the group of treated patients was 33 months. All had an outcome of light perception or no light perception visual acuity but had a comfortable eye, with none developing NVG over the mean follow-up of 28 months from presentation (median, 20 months). Among the control group of 6 eyes that was observed without any treatment, 4 were eventually enucleated for a blind, painful eye, indicating the natural history of advanced Coats disease. This seminal study revolutionized the management of advanced Coats disease with a shift away from enucleation toward surgical interventions to preserve the eye when the diagnosis of stage 3B Coats disease is confirmed. However, although treated eyes seemed to do well in the first 2 years of follow-up, the subsequent course was unknown. We report the long-term outcomes of a multicenter review of patients with treated stage 3B Coats disease, each followed for at least 5 years.
Methods Approval for this retrospective series was obtained from the institutional review boards at each of the participating sites. The research adhered to the tenets of the Declaration of Helsinki. All patients presenting with stage 3B Coats disease (total ERD) were reviewed, noting presentation, interventions, course, final visual and anatomic outcome, and length of follow-up. Retinoblastoma was ruled out. Patients with less than 5 years of total follow-up were excluded because there were insufficient data to draw any conclusions about long-term outcomes after treatment; of note, none of these excluded patients were lost to follow-up before 5 years or had a poor outcome (e.g., loss of light perception or enucleation). Treatment was performed by 5 pediatric retina surgeons (A.C., M.T.T., J.E.S., A.K., P.J.F.) over the entire study period with informed consent from each patient or his or her parent or guardian. All cases received ablative treatment, consisting of 1 or more sessions of cryotherapy or photocoagulation. Further treatment was given according to the judgment and preference of each individual surgeon.
Results There were 16 patients identified with stage 3B Coats disease followed for at least 5 years. The overall mean follow-up was 115 months (median, 114 months), ranging from 60 to 226 months. Their courses are summarized in Tables 1 to 3. Patients presented from 9 to 170 months of age (mean, 53 months; median, 41 months). A total of 11 of 16 were delivered as full-term babies; birth history was not known for the other 5
2
patients. As expected, most were male (13/16); of note, 1 of the female babies (case 4) had a history of Turner syndrome, and had various conditions known to be associated with this chromosomal anomaly, including coarctation of the aorta that was repaired, horseshoe kidney with congenital hydronephrosis, and kyphosis. All patients in this series had unilateral stage 3B Coats disease with total ERDs, 9 of 16 occurring in the right eye. On presentation, vision in the affected eye ranged from 20/100 to light perception. Two patients presented with neovascularization; 1 of them had neovascularization of the disc, and the other patient had both neovascularization of the disc and neovascularization elsewhere. All patients exhibited peripheral ischemia.
Management Patients were managed initially with several sessions of ablative treatment with laser photocoagulation or external cryotherapy. On the basis of surgeon preference, this initial ablative therapy was supplemented with additional interventions, which included, over the course of their management, pars plana vitrectomy (PPV) in 12 cases, scleral buckle in 8 (4 in the early vitrectomy [EV] group; 2 in the late vitrectomy group, and 2 in the no vitrectomy group), and XD in 6 (4 in the EV group, 1 in the late vitrectomy group, and 1 in the no vitrectomy group). External drainage of subretinal fluid was performed while reinsufflating the globe with an anterior chamber maintainer. No patients were drained internally. Eight patients underwent early PPV, performed soon after presentation (arbitrarily decided to be within 1 year); the mean time to vitrectomy in this group undergoing early PPV was 5 months (range, 0e12 months) (Table 1). In the remaining 8 cases, PPV was performed when the treating pediatric retina surgeon judged it clinically warranted. Four cases underwent late PPV, all for treatment of progressive clinically significant traction retinal detachment (TRD); the mean time to vitrectomy in the group undergoing late PPV was 62 months (range, 33e139 months) (Table 2). The other 4 patients did not require PPV over the course of long-term follow-up because they did not develop any significant vitreoretinal traction (Table 3). Intravitreal anti-vascular endothelial growth factor (VEGF) agents were used in only 2 patients in this series. In case 6, bevacizumab was used late in the course during progression to NVG in stage 4 Coats disease for the management of anterior segment neovascularization (Table 1). Case 16 received pegaptanib at the time of vitrectomy and received bevacizumab 2 years postvitrectomy in an attempt to further treat a recurrence of exudation (Table 1).
Vitreoretinal Fibrosis and Traction Retinal Detachments Six of 16 patients developed vitreoretinal fibrosis; 5 were progressive TRDs, and 1 (case 16) developed epiretinal membranes. All 6 cases were treated with PPV. In case 16, the patient presented at age 10 years with stage 3B Coats disease, and early PPV was performed in conjunction with lensectomy, photocoagulation, and pegaptanib at 7 months after presentation, after inadequate response to 3 sessions of cryotherapy initially. This patient ultimately required another anti-VEGF treatment with bevacizumab approximately 2 years after PPV for recurrent exudation with epiretinal membranes noted at final follow-up approximately 10 years after PPV. In the 5 patients who developed progressive TRDs, PPV was not performed until the TRD became clinically significant and posed a risk for developing a retinal tear from the vitreoretinal traction. Initially, these 5 patients (cases 1e5) reattached after
No
No
No
No
No
No No Unknown 96 9
20/400
72
HM
Cryotherapy
No Unknown 72 10
20/100
153
LP
Cryotherapy
Yes 20/200 121 16
20/200
127
20/200 Cryotherapy (3)
Yes CF HM 59 HM 170 2
CF ¼ counting fingers; EL ¼ endolaser; HM ¼ hand motions; LP ¼ light perception; NLP ¼ no light perception; NVG ¼ neovascular glaucoma; PCL ¼ posterior chamber lens; PFO ¼ perfluorocarbon liquid; PPL ¼ pars plana lensectomy; PPV ¼ pars plana vitrectomy; SB ¼ scleral buckling; SOP ¼ silicone oil placement; TRD ¼ traction retinal detachment; XD ¼ external drainage of subretinal fluid.
No No
TRD with subretinal fibrosis and vitreous membranes PPV/PPL/EL/pegaptanib (7) Attached with macular and peripheral exudates, epiretinal membrane, hemorrhages and RPE changes Cryotherapy/SB followed by Stable serous retinal detachment phaco/PCL/20 g PPV/EL/SOP (9) with no active Coats lesions. No pain. SB/25 g PPV/EL/cryotherapy/SOP (12) Persistent subretinal fluid
Yes Yes PPV/SOP (2) No 25 12
LP
104
LP
LP
25 g PPV (7)
No No No No No No Unknown HM 11 9 7 8
LP HM
139 125
LP HM
Cryotherapy (2) SB/20 g PPV/PPL/PFO-assisted XD/EL/SOP Cryotherapy (1), photocoagulation (2), SB/XD/photocoagulation Photocoagulation (3)
PPV/PPL/XD (0) See initial treatment (0)
NVG/phthisis
Yes Yes No Unknown NLP 138 HM 35
Final Retinal Outcome
Attached postvitrectomy with late progression to phthisis Attached Attached after silicone oil removal
Long-Term Outcomes in Stage 3B Coats Disease
Cryotherapy (2)
PPV/XD (0)
6
Vitrectomy (mos after Presentation) Postablation Visual Acuity TRD Initial Ablation Treatment(s) (No.) Initial Total Final Presenting Visual Follow-up Visual Case Age (mos) Acuity (mos) Acuity
Table 1. Clinical Course of Patients with Stage 3B Coats Disease Treated with Early Vitrectomy (within 1 Year of Presentation)
NVG Phthisis
Li et al
ablative treatment with improved postablation visual acuities in 4 of 5 patients. Despite these initial improvements in vision, all 5 patients went on to develop progressive TRDs with a mean onset of 20 months after initial ablation (median, 23 months; range, 4e27 months) (Table 4). The location of the TRD was peripheral in 4 patients and posterior in 1 patient. The TRDs became clinically significant at a mean of 2.6 years after they were initially detected (median, 10 months; range, 3 months to 9.3 years), at which point they underwent 25-gauge PPV.
Final Anatomic and Visual Outcomes At final follow-up, 75% (12/16) of the stage 3B Coats disease cases in this series demonstrated no progression to stage 4 or stage 5, namely, NVG or phthisis bulbi, respectively. Four cases progressed to NVG or phthisis or both; 2 had early PPV (cases 6 and 12), and 2 had late PPV (cases 4 and 5). Among these cases that progressed, only 1 required enucleation for pain (case 4). A delayed-onset pigmentary retinopathy with vessel attenuation developed in 4 patients (cases 3, 4, 5, 15). Three of the 4 lost light perception. They did not develop any active Coats lesions as they lost vision. The other patient, case 3, has maintained hand motion visual acuity. The last examination of case 15 demonstrated subretinal hemorrhage and cholesterol crystals and atrophy, but no progression to NVG or phthisis. After the development of this pigmentary retinopathy and losing light perception, cases 4 and 5 subsequently progressed to phthisis with posterior segment disorganization by the end of follow-up, with case 5 requiring enucleation for pain. In terms of visual outcomes at final follow-up, 4 of 16 were no light perception, 4 of 16 were light perception, 4 of 16 were hand motion, 2 of 16 were counting fingers, and 2 of 16 had vision greater than or equal to 20/400. A total of 8 of 16 (50%) had light perception or no light perception visual acuity at the end of 9.5 years mean follow-up. This proportion was consistent among groups regardless of timing of PPV, with 4 of 8 who had early PPV, 2 of 4 who had late PPV, and 2 of 4 who had no PPV ending with light perception or no light perception visual acuity.
Discussion In this series of 16 cases of treated stage 3B Coats disease, 50% had visual acuities of light perception or no light perception at the end of a mean 115 months (9 1/2 years) of follow-up, with 25% of overall patients advancing to NVG or phthisis or both. One of these patients who progressed required enucleation. This progression rate to more advanced Coats disease is worse than the 0% progression rate and 0% requiring enucleation among the treated patients in the series by Silodor et al4 with 28 months (21/3 years) of mean follow-up. This suggests that although patients seem to do well with interventions and avoid progression in the first 2 years of follow-up, further follow-up past 5 years shows some patients have progression to NVG and phthisis despite these interventions. Our overall observation of the course of stage 3B Coats disease has been that although vision is maintained or even improved in the first couple of years with initial ablation and interventions, later in the course, the vision can worsen to the point of no light perception, in some cases associated with a progressive pigmentary retinopathy. After the loss of retinal function, the posterior segment subsequently can undergo disorganization with development of vitreous hemorrhage, rubeosis
3
25 g PPV/SB (139) Yes 20/400 Photocoagulation (2) CF 226 29 1
LP
25 g PPV/SB (41) Yes 20/300 Photocoagulation (2); cryotherapy (1) NLP (before redetachment) 91 47 5
20/400
25 g PPV (36) Yes 20/300 NLP 106 LP 22 4
CF ¼ counting fingers; HM ¼ hand motions; LP ¼ light perception; NLP ¼ no light perception; NVG ¼ neovascular glaucoma; PPV ¼ pars plana vitrectomy; SB ¼ scleral buckling; TRD ¼ traction retinal detachment.
No No
No
Yes (requiring enucleation for pain) Yes Yes
No No
NVG Final Retinal Outcome
Attached with diffuse retinal pigmentary changes but no active Coats lesions Posterior segment disorganization (vitreous hemorrhage, retropupillary membrane, rubeosis iridis) Posterior segment disorganization (limited traction and rhegmatogenous redetachment with stage 3 macular hole; hyphema, vitreous hemorrhage) Attached with macular scar, preretinal and subretinal fibrosis, peripheral sclerotic vessels but no active Coats lesions 25 g PPV (33) Yes LP
Photocoagulation (3); XD (1); cryotherapy (1) Photocoagulation (5); cryotherapy (4) HM 137 LP
TRD Case
11
Vitrectomy (mos after Presentation) Post-ablation Visual Acuity Initial Ablation Treatment(s) (No.) Final Visual Acuity Total Follow-up (mos) Presenting Age (mos)
Initial Visual Acuity
Table 2. Clinical Course of Patients with Stage 3B Coats Disease Treated with Late Vitrectomy (>1 Year after Presentation)
4
3
Phthisis
Ophthalmology Volume -, Number -, Month 2017 iridis, NVG, and ultimately phthisis bulbi. Subgroup analysis does not point to 1 set of interventions being more effective in preventing loss of vision or progression to NVG or phthisis. However, given that there were fewer instances of vitreoretinal traction in cases undergoing EV, vitrectomy sooner rather than later may help prevent the development of TRDs. Although this series had better final visual acuity outcomes than the 100% of treated patients who had no light perception or light perception in the series by Silodor et al,4 all of the eyes in the series by Silodor et al presented with no fixing or following vision, whereas 4 of 16 in our series had visual acuities greater than or equal to 20/400 on presentation. Another confounder when comparing visual acuities is the difficulty in obtaining consistent visual acuities in extremely young patients, because the series by Silodor et al4 only recorded “fix and follow” vision. This series does concur with the series by Silodor et al4 in supporting the role of ablative therapy and surgical intervention, which results in an improved outcome over the natural history of advanced Coats diseasedenucleation for a blind, painful eye within 1 year of initial presentation with stage 3B disease.4 Although Silodor et al4 exclusively reported on the use of XD, this series expands the armamentarium to include scleral buckling (SB) and vitrectomy, which has undoubtedly benefited from the advances in instrumentation over the last few decades. Aggressive ablative treatment with a combination of possible vitrectomy, SB, or XD if necessary, can help preserve a cosmetically acceptable and comfortable globe. In the worst case, it delays enucleation by several more years. In this series of 16 eyes, enucleation was required in 1 case. Even if enucleation eventually becomes necessary, the delay to enucleation allows time for the orbit to mature further and aids in better fitting of an orbital implant. We hypothesize that the benefit of XD encountered both in the series by Silodor et al4 and our series may be related to at least 2 factors: (1) clearing the toxic milieu in which the photoreceptors are bathed; and (2) eliminating proinflammatory cytokines and profibrotic signals, such as VEGF. Given that there were fewer instances of vitreoretinal traction in patients undergoing EV, vitrectomy may help prevent the development of TRDs in a similar manner by clearing profibrotic signals from the vitreous and of course by eliminating the transvitreal collagen framework, which facilitates the formation of these TRDs. If this is true, it would follow that there would be an added advantage when XD and vitrectomy are performed earlier in the disease course. However, we are not recommending vitrectomy summarily as the first intervention for stage 3B Coats disease, because a majority of those in the EV group did not have vitrectomy until several ablative treatments were completed; ablation of the telangiectatic vessels in Coats disease remains a fundamental principle for treatment, regardless of stage. This series also showed that the onset of TRDs was late and progressive, most commonly starting in the periphery. Scleral buckling generates retinal redundancy that would facilitate maintaining retinal attachment in light of
Li et al
Long-Term Outcomes in Stage 3B Coats Disease
Table 3. Clinical Course of Patients with Stage 3B Coats Disease Treated with No Vitrectomy (>1 Year after Presentation) Total Presenting Initial Visual Follow-up Final Visual Case Age (mos) Acuity (mos) Acuity 11 13
39 42
CF CF 2’
120 72
CF 20/400
14
54
LP
60
LP
15
57
LP
108
NLP
Initial Ablation Treatment(s) (No.) SB/cryotherapy Photocoagulation, cryotherapy SB/cryotherapy/XD, photocoagulation (4) Cryotherapy (3), photocoagulation (1)
Post-ablation Visual Acuity TRD
Final Retinal Outcome
NVG Phthisis
CF CF
No No
Attached, macular scar Attached, macular scar
No No
No No
LP
No
Attached
No
No
LP
No
Subretinal hemorrhage and cholesterol crystals, pigmentary changes, retinal atrophy
No
No
CF ¼ counting fingers; LP ¼ light perception; NLP ¼ no light perception; NVG ¼ neovascular glaucoma; SB ¼ scleral buckling; TRD ¼ traction retinal detachment; XD ¼ external drainage of subretinal fluid.
development of this traction, which would provide a theoretic protective benefit in the later cicatricial phase of stage 3B Coats disease, when vitreoretinal traction begins to form. This is all balanced by the fact that there were 4 of 16 patients who did not require any vitrectomy because they did not develop any significant vitreoretinal traction at any point. Of note, 2 of these 4 patients did receive an SB early in their treatment regimen. It is not clear if this had any effect on the lack of need for a vitrectomy later. In this age of intravitreal injections, anti-VEGF agents and triamcinolone have emerged as adjuncts to the traditional methods of ablative treatment; anti-VEGF therapy has been shown to decrease macular edema and exudates5 and even reverse a total exudative detachment.6 However, the use of intravitreal bevacizumab and ranibizumab in Coats disease was thought to be possibly associated with the development of vitreoretinal fibrosis and TRD.7,8 This adverse effect has also been reported in the context of antiVEGF treatment of other ischemic retinopathies.8 In our series, 6 of 16 patients developed vitreoretinal fibrosis, of whom 5 developed TRDs. In only 1 of these patients was anti-VEGF treatment given, and it was given at the time of vitrectomy with the patient developing epiretinal membranes at the time of his last examination, seemingly unrelated to the anti-VEGF use. In the other 5 patients to whom no anti-VEGF was given, vitrectomy was performed when the TRDs became clinically significant at a mean of 2.6 years after onset of TRD, which in turn occurred a mean 20 months after initial ablative treatment of the ERD. According to the Shields classification, Coats disease is not characterized by vitreoretinal fibrosis, yet the late development of TRDs in the absence of anti-VEGF in these 5 patients suggests that this could represent a feature of advanced Coats disease. Vitreoretinal fibrosis and subsequent TRD have been reported earlier in the course of advanced Coats disease subsequent to intravitreal injection of bevacizumab8 and ranibizumab.7 Ramasubramanian and Shields8 reported 8 eyes with stage 2 to 3B Coats disease treated with intravitreal bevacizumab at the same time as standard treatment of laser photocoagulation, cryotherapy, and subTenon’s triamcinolone; 50% (4/8) developed vitreous
fibrosis at a mean of 5 months post-treatment, and 3 of the 4 progressed to TRD.8 Gaillard et al7 presented 5 cases of stage 3B Coats disease and 4 cases of stage 4 Coats disease treated with subretinal drainage to reform the vitreous cavity, with some patients also having intravitreal ranibizumab and subsequent treatment with photocoagulation and cryocoagulation. Overall, their mean follow-up was 46 months, with 5 of 9 developing vitreoretinal fibrosis, and 1 of whom developed a TRD followed by phthisis bulbi.7 In the subset of stage 3B Coats disease in the series by Gaillard et al,7 40% (2/5) developed vitreoretinal fibrosis at a mean of 15 months after ranibizumab injection. However, the mean follow-up for their subset of stage 3B patients was 40 months, with only 3 of these 5 followed for more than 3 years. In comparison, all 16 patients in our series had at least 5 years of follow-up. Of the 6 cases with vitreoretinal fibrosis in our series, 1 had antiVEGF use; however, the formation of epiretinal membranes in this patient was noted at the final follow-up 10 years later. Given this timing, it is unlikely that anti-VEGF played a role in the formation of epiretinal membranes in this case. The other 5 cases had progressive vitreous organization with eventual TRD after initial management Table 4. Timing of Onset of Vitreoretinal Fibrosis and Traction Retinal Detachment in Patients with Stage 3B Coats Disease Who Developed Vitreoretinal Fibrosis Time from Resolution Time from Onset of Active Lesions of TRD to Clinically Time from after Ablation to Significant TRD and Presentation to Case Onset of TRD (Mos) Vitrectomy (Mos) Vitrectomy (Mos) 2 16 3 4 5 1
4 N/A* 23 27 17 27
3 N/A 10 9 24 112
7 7 33 36 41 139
N/A ¼ not available; TRD ¼ traction retinal detachment. *Vitreoretinal fibrosis in the form of epiretinal membranes occurred late (10 years) after vitrectomy. See text for details.
5
Ophthalmology Volume -, Number -, Month 2017 without vitrectomy, with a mean 20 months from presentation to initial signs of vitreoretinal fibrosis and another mean 2.4 years for fibrosis to progress to a clinically significant TRD requiring vitrectomy. Subretinal and vitreoretinal fibrosis in Coats disease, whether in the macula or outside of the macula, is associated with poorer visual outcomes with the prevalence estimated to be 23%9 and 41%,10 respectively. Daruich et al10 reported 69 cases of Coats disease, ranging in severity from stage 1 to 4; their treatment protocol included laser photocoagulation targeting the peripheral telangiectasia, cryotherapy, subretinal fluid drainage, and intravitreal anti-VEGF in cases of severe ERD. They noted that 41% (28/69) of eyes with Coats disease developed extramacular fibrosis at a mean of 17 months after diagnosis (mean follow-up, 80 months). The onset of fibrosis noted by Daruich et al10 is close to the onset noted by Gaillard et al7 of 15 months and our finding of the onset of TRD at 20 months after ablative treatment. These instances of extramacular fibrosis can be further classified as vitreoretinal fibrosis alone (6/28), subretinal fibrosis alone (2/28), and vitreoretinal fibrosis with subretinal fibrosis (20/28). Among these 28 patients with extramacular fibrosis, 11 of 28 developed TRDs, which they found to be associated with severe vision loss; 63% (7/11) of eyes with TRD presented with no light perception at their last follow-up, compared with 3 of 58 who had no TRD had no light perception in their study.10 The pathogenesis of fibrosis in Coats disease has yet to be fully elucidated, with some suggesting that it could represent the natural course of the disease, given that they encountered vitreoretinal fibrosis before any treatment, antiVEGF or otherwise.11 In the study by Daruich et al,10 the number of patients receiving intravitreal bevacizumab or ranibizumab did not differ significantly between patients who developed extramacular subretinal or vitreoretinal fibrosis and those who did not. In our study, 5 of 6 patients developing vitreoretinal traction did not receive anti-VEGF therapy. In fact, overall anti-VEGF was used sparingly, ultimately in the late management of 2 patients. Its use in 1 case (case 6) was for the management of progression to stage 4 Coats disease to control anterior segment neovascularization and thus not relevant to the discussion of pathogenesis of vitreoretinal traction. In the other case (case 16), the patient received anti-VEGF treatment twice and developed epiretinal membranes at the end of follow-up. Given the incidence of vitreoretinal fibrosis in the absence of anti-VEGF use in 5 cases in this series, this supports the idea that vitreoretinal fibrosis and tractional detachment represents the end-stage in the natural history of exudative Coats disease as posited by Zheng and Jiang,11 as opposed to an adverse effect of anti-VEGF treatment as proposed by Ramasubramanian and Shields.8 In analyzing the various factors that contribute to extramacular subretinal or vitreoretinal fibrosis, Daruich et al10 also showed that the type of ERD, whether total, partial, or none, and the degree of retinal exudation, measured by clock-hours, were significant independent contributing factors to the development of extramacular subretinal or vitreoretinal fibrosis. Furthermore, the extent of retinal exudation, a defining characteristic of Coats disease, and
6
degree of ERD were shown to be significantly higher in eyes with extramacular subretinal or vitreoretinal fibrosis.10 It is evident from the literature that both subretinal fibrosis and vitreoretinal fibrosis are linked to poor outcomes. A few case reports have measured a high level of VEGF in the vitreous, subretinal fluid, and aqueous of eyes with Coats disease, which subsequently decreased after intravitreal anti-VEGF treatment along with some clinical improvement.12,13 These reports suggest a role for the dysregulation of VEGF in the pathogenesis of Coats disease and a rationale for anti-VEGF agents in the treatment of Coats disease. Nevertheless, on the basis of the experience of Ramasubramanian and Shields,8 regimens involving antiVEGF in Coats disease may need to be optimized with the other modalities to maximize vision and globe salvage. In terms of timing of vitrectomy with respect to the incidence of vitreoretinal traction, 2 of 8 patients who underwent early PPV (within the first year of presentation) developed vitreoretinal traction, whereas 4 of 8 in the expectant management (late PPV or no PPV) had developed vitreoretinal traction. In the 2 cases that underwent early PPV, 1 was for the treatment of TRD that developed at 7 months after presentation, whereas the other had EV and developed epiretinal membranes in a delayed fashion approximately a decade after PPV. The other 6 patients who underwent early PPV did not show signs of subsequent vitreoretinal fibrosis. In the group of patients in whom PPV was delayed until judged to be necessary, 4 of 8 required PPV for progressive vitreoretinal traction that developed into a late-onset clinically significant TRD. Although the numbers were small, we did see that there were fewer cases of vitreoretinal traction in the group that underwent early (prophylactic) vitrectomy. However, a conclusion of benefit cannot be firmly reached because of the nature of this retrospective study and has to be balanced by the fact that 4 of 8 of the expectant management group never required any vitrectomy for vitreoretinal traction. As expected, most of the patients in our series were male. However, the occurrence of advanced Coats disease in a girl with Turner syndrome (case 4) merits further consideration. Although Coats disease is not a common finding in Turner syndrome, cases of patients with both conditions have been reported in the literature.14e16 Once thought to be a “chance occurrence of the 2 syndromes,”14 the presence of Coats disease in a patient with Turner syndrome could represent another vascular anomaly associated with Turner syndrome, such as aortic coarctation, cerebral vessel aneurysm, and gastrointestinal telangiectasia.15 It may be that additional investigation into the pathogenesis of these vascular anomalies in Turner syndrome could advance our understanding of Coats disease. The limitations of this study include its retrospective nature and that treatment was not standardized across surgeons. Nevertheless, it is significant in that it presents longterm follow-up in stage 3B Coats disease. Overall, the globe was salvaged in 15 of 16 cases with a mean follow-up of 9.5 years. These interventions with long-term follow-up could not reverse the poor visual outcomes associated with advanced Coats disease. These treated eyes do better than the natural history of stage 3B Coats disease. In the majority
Li et al
Long-Term Outcomes in Stage 3B Coats Disease
of cases, treatment was successful in preserving the globe without pain. In the patient in whom enucleation became necessary, these interventions allowed the globe to remain for several more years than otherwise expected without treatment and allowed the orbit to mature further to better accommodate a prosthesis in the future. Vitrectomy at the time of initial treatment may be beneficial in preventing delayed onset vitreous organization and TRD, although a larger prospective study comparing treatments looking at the development of fibrosis in Coats disease is necessary to direct management and possible timing of vitrectomy. Furthermore, a better understanding of fibrosis in the progression of advanced Coats disease would be helpful in guiding management and possibly developing improved interventions.
References 1. Shields JA, Shields CL, Honavar SG, et al. Classification and management of Coats disease: the 2000 Proctor Lecture. Am J Ophthalmol. 2001;131:572e583. 2. Shields JA, Shields CL, Honavar SG, Demirci H. Clinical variations and complications of Coats disease in 150 cases: the 2000 Sanford Gifford Memorial Lecture. Am J Ophthalmol. 2001;131:561e571. 3. Ghorbanian S, Jaulim A, Chatziralli IP. Diagnosis and treatment of Coats’ disease: a review of the literature. Ophthalmologica. 2012;227:175e182. 4. Silodor SW, Augsburger JJ, Shields JA, Tasman W. Natural history and management of advanced Coats’ disease. Ophthalmic Surg. 1988;19:89e93. 5. Sigler EJ, Randolph JC, Calzada JI, et al. Current management of Coats disease. Surv Ophthalmol. 2014;59:30e46.
6. Zhao T, Wang K, Ma Y, Jiang YR. Resolution of total retinal detachment in Coats’ disease with intravitreal injection of bevacizumab. Graefes Arch Clin Exp Ophthalmol. 2011;249: 1745e1746. 7. Gaillard MC, Mataftsi A, Balmer A, et al. Ranibizumab in the management of advanced Coats disease Stages 3B and 4: longterm outcomes. Retina. 2014;34:2275e2281. 8. Ramasubramanian A, Shields CL. Bevacizumab for Coats’ disease with exudative retinal detachment and risk of vitreoretinal traction. Br J Ophthalmol. 2012;96: 356e359. 9. Jumper JM, Pomerleau D, McDonald HR, et al. Macular fibrosis in Coats disease. Retina. 2010;30:S9e14. 10. Daruich A, Matet A, Tran HV, et al. Extramacular fibrosis in Coats’ disease. Retina. 2016;36:2022e2028. 11. Zheng XX, Jiang YR. The effect of intravitreal bevacizumab injection as the initial treatment for Coats’ disease. Graefes Arch Clin Exp Ophthalmol. 2014;252:35e42. 12. He YG, Wang H, Zhao B, et al. Elevated vascular endothelial growth factor level in Coats’ disease and possible therapeutic role of bevacizumab. Graefes Arch Clin Exp Ophthalmol. 2010;248:1519e1521. 13. Sun Y, Jain A, Moshfeghi DM. Elevated vascular endothelial growth factor levels in Coats disease: rapid response to pegaptanib sodium. Graefes Arch Clin Exp Ophthalmol. 2007;245:1387e1388. 14. Cameron JD, Yanoff M, Frayer WC. Coats’ disease and Turner’s syndrome. Am J Ophthalmol. 1974;78:852e854. 15. Beby F, Roche O, Burillon C, Denis P. Coats’ disease and bilateral cataract in a child with Turner syndrome: a case report. Graefes Arch Clin Exp Ophthalmol. 2005;243: 1291e1293. 16. Desai RU, Saffra NA, Krishna RP, Rosenberg SE. Coats’ disease, Turner syndrome, and von Willebrand disease in a patient with Wildtype Norrie disease pseudoglioma. J Pediatr Ophthalmol Strabismus. 2011;48 Online:e1e3.
Footnotes and Financial Disclosures Originally received: May 18, 2017. Final revision: November 30, 2017. Accepted: December 5, 2017. Available online: ---. 1 Hofstra Northwell School of Medicine, Great Neck, New York. 2
Associated Retinal Consultants, Royal Oak, Michigan.
3
Cleveland Clinic, Cleveland, Ohio.
4
KYDOFT Foundation, Santiago, Chile.
5
Long Island Vitreoretinal Consultants, Great Neck, New York.
Presented as a poster at: the American Academy of Ophthalmology Annual Meeting, November 11e14, 2017, New Orleans, Louisiana. Financial Disclosure(s): The author(s) have made the following disclosure(s): P.J.F.: Consultant Genentech, Allergan, Alcon, and Santeen. HUMAN SUBJECTS: This study includes human subject/tissues. Study protocol was approved by IRB of North Shore University Hospital. All tenets of the Declaration of Helsinki were followed.
Author Contributions: Conception and design: Li, Ferrone Data collection: Li, Capone Jr., Trese, Sears, Kychenthal, De la Huerta, Ferrone Analysis and interpretation: Li, Ferrone Obtained funding: Not applicable Overall responsibility: Li, Capone Jr., Trese, Sears, Kychenthal, De la Huerta, Ferrone Abbreviations and Acronyms: ERD ¼ exudative retinal detachment; EV ¼ early vitrectomy; NVG ¼ neovascular glaucoma; PPV ¼ pars plana vitrectomy; TRD ¼ traction retinal detachment; VEGF ¼ vascular endothelial growth factor; XD ¼ external drainage of subretinal fluid. Correspondence: Philip J. Ferrone, MD, Long Island Vitreoretinal Consultants, 600 Northern Boulevard, Suite 216, Great Neck, NY 11021. E-mail: PJFerrone@gmail. com.
7
Coats disease is a nonhereditary disorder of the retinal vasculature, characterized by leaky retinal telangiectasia and associated subretinal or intraretinal exudation.1 It usually is present unilaterally and predominantly affects young men. There is a wide spectrum of presentation for Coats disease; it can be found as an asymptomatic retinal telangiectasia noted on examination or can present with exuberant exudation causing retinal detachment, progressing to secondary neovascular glaucoma (NVG) and phthisis. Shields et al2 proposed the definition of Coats disease as an “idiopathic retinal telangiectasia with intraretinal and/or subretinal exudation without appreciable retinal or vitreal traction.” Drawing on their experience with 150 such cases, they also presented a staging system that classifies Coats disease on the basis of the severity of these findings.2 Management of Coats disease varies according to the stage of disease. In cases of retinal telangiectasia only (stage 1) or telangiectasias with exudation (stage 2) that is ª 2017 by the American Academy of Ophthalmology Published by Elsevier Inc.
not vision threatening, close observation can be used. Laser photocoagulation and cryotherapy constitute the mainstay treatment for Coats disease once exudation becomes significant. Both are options in cases of exudative subtotal retinal detachment (stage 3A) and even exudative total retinal detachment (stage 3B).3 Nonetheless, treatment of stage 3A and 3B Coats disease is challenging because the exuberant exudation characteristic of this stage is not as responsive to ablative therapy and often requires multiple sessions to eliminate all active and recurrent Coats lesions. Stage 3 Coats disease is also on the cusp of progressing onto intractable disease, with stage 4 characterized by NVG and stage 5 marked by a blind eye with cataract and phthisis bulbi. For stage 3B, the goals of treatment are maintaining any residual vision and globe preservation without pain. Advanced Coats disease with prolific exudation can present with leukocoria and be difficult to distinguish from retinoblastoma on examination. In such cases in which https://doi.org/10.1016/j.ophtha.2017.12.010 ISSN 0161-6420/17
1
Ophthalmology Volume -, Number -, Month 2017 retinoblastoma cannot be definitively excluded, ophthalmologists have opted to enucleate an eye with poor prognosis so as not to delay diagnosis and definitive treatment of a potentially lethal condition, inevitably also enucleating some eyes with advanced Coats disease. However, in 1988, Silodor et al4 showed that surgical intervention in eyes with confirmed stage 3B Coats disease was effective in salvaging the globe for cosmesis and staving off progression to painful NVG. In their series of 13 patients with total bullous exudative retinal detachments (ERDs) from advanced unilateral Coats disease, 7 eyes were treated with intraocular infusion, external drainage of subretinal fluid (XD) and cryotherapy; the mean age at diagnosis of the group of treated patients was 33 months. All had an outcome of light perception or no light perception visual acuity but had a comfortable eye, with none developing NVG over the mean follow-up of 28 months from presentation (median, 20 months). Among the control group of 6 eyes that was observed without any treatment, 4 were eventually enucleated for a blind, painful eye, indicating the natural history of advanced Coats disease. This seminal study revolutionized the management of advanced Coats disease with a shift away from enucleation toward surgical interventions to preserve the eye when the diagnosis of stage 3B Coats disease is confirmed. However, although treated eyes seemed to do well in the first 2 years of follow-up, the subsequent course was unknown. We report the long-term outcomes of a multicenter review of patients with treated stage 3B Coats disease, each followed for at least 5 years.
Methods Approval for this retrospective series was obtained from the institutional review boards at each of the participating sites. The research adhered to the tenets of the Declaration of Helsinki. All patients presenting with stage 3B Coats disease (total ERD) were reviewed, noting presentation, interventions, course, final visual and anatomic outcome, and length of follow-up. Retinoblastoma was ruled out. Patients with less than 5 years of total follow-up were excluded because there were insufficient data to draw any conclusions about long-term outcomes after treatment; of note, none of these excluded patients were lost to follow-up before 5 years or had a poor outcome (e.g., loss of light perception or enucleation). Treatment was performed by 5 pediatric retina surgeons (A.C., M.T.T., J.E.S., A.K., P.J.F.) over the entire study period with informed consent from each patient or his or her parent or guardian. All cases received ablative treatment, consisting of 1 or more sessions of cryotherapy or photocoagulation. Further treatment was given according to the judgment and preference of each individual surgeon.
Results There were 16 patients identified with stage 3B Coats disease followed for at least 5 years. The overall mean follow-up was 115 months (median, 114 months), ranging from 60 to 226 months. Their courses are summarized in Tables 1 to 3. Patients presented from 9 to 170 months of age (mean, 53 months; median, 41 months). A total of 11 of 16 were delivered as full-term babies; birth history was not known for the other 5
2
patients. As expected, most were male (13/16); of note, 1 of the female babies (case 4) had a history of Turner syndrome, and had various conditions known to be associated with this chromosomal anomaly, including coarctation of the aorta that was repaired, horseshoe kidney with congenital hydronephrosis, and kyphosis. All patients in this series had unilateral stage 3B Coats disease with total ERDs, 9 of 16 occurring in the right eye. On presentation, vision in the affected eye ranged from 20/100 to light perception. Two patients presented with neovascularization; 1 of them had neovascularization of the disc, and the other patient had both neovascularization of the disc and neovascularization elsewhere. All patients exhibited peripheral ischemia.
Management Patients were managed initially with several sessions of ablative treatment with laser photocoagulation or external cryotherapy. On the basis of surgeon preference, this initial ablative therapy was supplemented with additional interventions, which included, over the course of their management, pars plana vitrectomy (PPV) in 12 cases, scleral buckle in 8 (4 in the early vitrectomy [EV] group; 2 in the late vitrectomy group, and 2 in the no vitrectomy group), and XD in 6 (4 in the EV group, 1 in the late vitrectomy group, and 1 in the no vitrectomy group). External drainage of subretinal fluid was performed while reinsufflating the globe with an anterior chamber maintainer. No patients were drained internally. Eight patients underwent early PPV, performed soon after presentation (arbitrarily decided to be within 1 year); the mean time to vitrectomy in this group undergoing early PPV was 5 months (range, 0e12 months) (Table 1). In the remaining 8 cases, PPV was performed when the treating pediatric retina surgeon judged it clinically warranted. Four cases underwent late PPV, all for treatment of progressive clinically significant traction retinal detachment (TRD); the mean time to vitrectomy in the group undergoing late PPV was 62 months (range, 33e139 months) (Table 2). The other 4 patients did not require PPV over the course of long-term follow-up because they did not develop any significant vitreoretinal traction (Table 3). Intravitreal anti-vascular endothelial growth factor (VEGF) agents were used in only 2 patients in this series. In case 6, bevacizumab was used late in the course during progression to NVG in stage 4 Coats disease for the management of anterior segment neovascularization (Table 1). Case 16 received pegaptanib at the time of vitrectomy and received bevacizumab 2 years postvitrectomy in an attempt to further treat a recurrence of exudation (Table 1).
Vitreoretinal Fibrosis and Traction Retinal Detachments Six of 16 patients developed vitreoretinal fibrosis; 5 were progressive TRDs, and 1 (case 16) developed epiretinal membranes. All 6 cases were treated with PPV. In case 16, the patient presented at age 10 years with stage 3B Coats disease, and early PPV was performed in conjunction with lensectomy, photocoagulation, and pegaptanib at 7 months after presentation, after inadequate response to 3 sessions of cryotherapy initially. This patient ultimately required another anti-VEGF treatment with bevacizumab approximately 2 years after PPV for recurrent exudation with epiretinal membranes noted at final follow-up approximately 10 years after PPV. In the 5 patients who developed progressive TRDs, PPV was not performed until the TRD became clinically significant and posed a risk for developing a retinal tear from the vitreoretinal traction. Initially, these 5 patients (cases 1e5) reattached after
No
No
No
No
No
No No Unknown 96 9
20/400
72
HM
Cryotherapy
No Unknown 72 10
20/100
153
LP
Cryotherapy
Yes 20/200 121 16
20/200
127
20/200 Cryotherapy (3)
Yes CF HM 59 HM 170 2
CF ¼ counting fingers; EL ¼ endolaser; HM ¼ hand motions; LP ¼ light perception; NLP ¼ no light perception; NVG ¼ neovascular glaucoma; PCL ¼ posterior chamber lens; PFO ¼ perfluorocarbon liquid; PPL ¼ pars plana lensectomy; PPV ¼ pars plana vitrectomy; SB ¼ scleral buckling; SOP ¼ silicone oil placement; TRD ¼ traction retinal detachment; XD ¼ external drainage of subretinal fluid.
No No
TRD with subretinal fibrosis and vitreous membranes PPV/PPL/EL/pegaptanib (7) Attached with macular and peripheral exudates, epiretinal membrane, hemorrhages and RPE changes Cryotherapy/SB followed by Stable serous retinal detachment phaco/PCL/20 g PPV/EL/SOP (9) with no active Coats lesions. No pain. SB/25 g PPV/EL/cryotherapy/SOP (12) Persistent subretinal fluid
Yes Yes PPV/SOP (2) No 25 12
LP
104
LP
LP
25 g PPV (7)
No No No No No No Unknown HM 11 9 7 8
LP HM
139 125
LP HM
Cryotherapy (2) SB/20 g PPV/PPL/PFO-assisted XD/EL/SOP Cryotherapy (1), photocoagulation (2), SB/XD/photocoagulation Photocoagulation (3)
PPV/PPL/XD (0) See initial treatment (0)
NVG/phthisis
Yes Yes No Unknown NLP 138 HM 35
Final Retinal Outcome
Attached postvitrectomy with late progression to phthisis Attached Attached after silicone oil removal
Long-Term Outcomes in Stage 3B Coats Disease
Cryotherapy (2)
PPV/XD (0)
6
Vitrectomy (mos after Presentation) Postablation Visual Acuity TRD Initial Ablation Treatment(s) (No.) Initial Total Final Presenting Visual Follow-up Visual Case Age (mos) Acuity (mos) Acuity
Table 1. Clinical Course of Patients with Stage 3B Coats Disease Treated with Early Vitrectomy (within 1 Year of Presentation)
NVG Phthisis
Li et al
ablative treatment with improved postablation visual acuities in 4 of 5 patients. Despite these initial improvements in vision, all 5 patients went on to develop progressive TRDs with a mean onset of 20 months after initial ablation (median, 23 months; range, 4e27 months) (Table 4). The location of the TRD was peripheral in 4 patients and posterior in 1 patient. The TRDs became clinically significant at a mean of 2.6 years after they were initially detected (median, 10 months; range, 3 months to 9.3 years), at which point they underwent 25-gauge PPV.
Final Anatomic and Visual Outcomes At final follow-up, 75% (12/16) of the stage 3B Coats disease cases in this series demonstrated no progression to stage 4 or stage 5, namely, NVG or phthisis bulbi, respectively. Four cases progressed to NVG or phthisis or both; 2 had early PPV (cases 6 and 12), and 2 had late PPV (cases 4 and 5). Among these cases that progressed, only 1 required enucleation for pain (case 4). A delayed-onset pigmentary retinopathy with vessel attenuation developed in 4 patients (cases 3, 4, 5, 15). Three of the 4 lost light perception. They did not develop any active Coats lesions as they lost vision. The other patient, case 3, has maintained hand motion visual acuity. The last examination of case 15 demonstrated subretinal hemorrhage and cholesterol crystals and atrophy, but no progression to NVG or phthisis. After the development of this pigmentary retinopathy and losing light perception, cases 4 and 5 subsequently progressed to phthisis with posterior segment disorganization by the end of follow-up, with case 5 requiring enucleation for pain. In terms of visual outcomes at final follow-up, 4 of 16 were no light perception, 4 of 16 were light perception, 4 of 16 were hand motion, 2 of 16 were counting fingers, and 2 of 16 had vision greater than or equal to 20/400. A total of 8 of 16 (50%) had light perception or no light perception visual acuity at the end of 9.5 years mean follow-up. This proportion was consistent among groups regardless of timing of PPV, with 4 of 8 who had early PPV, 2 of 4 who had late PPV, and 2 of 4 who had no PPV ending with light perception or no light perception visual acuity.
Discussion In this series of 16 cases of treated stage 3B Coats disease, 50% had visual acuities of light perception or no light perception at the end of a mean 115 months (9 1/2 years) of follow-up, with 25% of overall patients advancing to NVG or phthisis or both. One of these patients who progressed required enucleation. This progression rate to more advanced Coats disease is worse than the 0% progression rate and 0% requiring enucleation among the treated patients in the series by Silodor et al4 with 28 months (21/3 years) of mean follow-up. This suggests that although patients seem to do well with interventions and avoid progression in the first 2 years of follow-up, further follow-up past 5 years shows some patients have progression to NVG and phthisis despite these interventions. Our overall observation of the course of stage 3B Coats disease has been that although vision is maintained or even improved in the first couple of years with initial ablation and interventions, later in the course, the vision can worsen to the point of no light perception, in some cases associated with a progressive pigmentary retinopathy. After the loss of retinal function, the posterior segment subsequently can undergo disorganization with development of vitreous hemorrhage, rubeosis
3
25 g PPV/SB (139) Yes 20/400 Photocoagulation (2) CF 226 29 1
LP
25 g PPV/SB (41) Yes 20/300 Photocoagulation (2); cryotherapy (1) NLP (before redetachment) 91 47 5
20/400
25 g PPV (36) Yes 20/300 NLP 106 LP 22 4
CF ¼ counting fingers; HM ¼ hand motions; LP ¼ light perception; NLP ¼ no light perception; NVG ¼ neovascular glaucoma; PPV ¼ pars plana vitrectomy; SB ¼ scleral buckling; TRD ¼ traction retinal detachment.
No No
No
Yes (requiring enucleation for pain) Yes Yes
No No
NVG Final Retinal Outcome
Attached with diffuse retinal pigmentary changes but no active Coats lesions Posterior segment disorganization (vitreous hemorrhage, retropupillary membrane, rubeosis iridis) Posterior segment disorganization (limited traction and rhegmatogenous redetachment with stage 3 macular hole; hyphema, vitreous hemorrhage) Attached with macular scar, preretinal and subretinal fibrosis, peripheral sclerotic vessels but no active Coats lesions 25 g PPV (33) Yes LP
Photocoagulation (3); XD (1); cryotherapy (1) Photocoagulation (5); cryotherapy (4) HM 137 LP
TRD Case
11
Vitrectomy (mos after Presentation) Post-ablation Visual Acuity Initial Ablation Treatment(s) (No.) Final Visual Acuity Total Follow-up (mos) Presenting Age (mos)
Initial Visual Acuity
Table 2. Clinical Course of Patients with Stage 3B Coats Disease Treated with Late Vitrectomy (>1 Year after Presentation)
4
3
Phthisis
Ophthalmology Volume -, Number -, Month 2017 iridis, NVG, and ultimately phthisis bulbi. Subgroup analysis does not point to 1 set of interventions being more effective in preventing loss of vision or progression to NVG or phthisis. However, given that there were fewer instances of vitreoretinal traction in cases undergoing EV, vitrectomy sooner rather than later may help prevent the development of TRDs. Although this series had better final visual acuity outcomes than the 100% of treated patients who had no light perception or light perception in the series by Silodor et al,4 all of the eyes in the series by Silodor et al presented with no fixing or following vision, whereas 4 of 16 in our series had visual acuities greater than or equal to 20/400 on presentation. Another confounder when comparing visual acuities is the difficulty in obtaining consistent visual acuities in extremely young patients, because the series by Silodor et al4 only recorded “fix and follow” vision. This series does concur with the series by Silodor et al4 in supporting the role of ablative therapy and surgical intervention, which results in an improved outcome over the natural history of advanced Coats diseasedenucleation for a blind, painful eye within 1 year of initial presentation with stage 3B disease.4 Although Silodor et al4 exclusively reported on the use of XD, this series expands the armamentarium to include scleral buckling (SB) and vitrectomy, which has undoubtedly benefited from the advances in instrumentation over the last few decades. Aggressive ablative treatment with a combination of possible vitrectomy, SB, or XD if necessary, can help preserve a cosmetically acceptable and comfortable globe. In the worst case, it delays enucleation by several more years. In this series of 16 eyes, enucleation was required in 1 case. Even if enucleation eventually becomes necessary, the delay to enucleation allows time for the orbit to mature further and aids in better fitting of an orbital implant. We hypothesize that the benefit of XD encountered both in the series by Silodor et al4 and our series may be related to at least 2 factors: (1) clearing the toxic milieu in which the photoreceptors are bathed; and (2) eliminating proinflammatory cytokines and profibrotic signals, such as VEGF. Given that there were fewer instances of vitreoretinal traction in patients undergoing EV, vitrectomy may help prevent the development of TRDs in a similar manner by clearing profibrotic signals from the vitreous and of course by eliminating the transvitreal collagen framework, which facilitates the formation of these TRDs. If this is true, it would follow that there would be an added advantage when XD and vitrectomy are performed earlier in the disease course. However, we are not recommending vitrectomy summarily as the first intervention for stage 3B Coats disease, because a majority of those in the EV group did not have vitrectomy until several ablative treatments were completed; ablation of the telangiectatic vessels in Coats disease remains a fundamental principle for treatment, regardless of stage. This series also showed that the onset of TRDs was late and progressive, most commonly starting in the periphery. Scleral buckling generates retinal redundancy that would facilitate maintaining retinal attachment in light of
Li et al
Long-Term Outcomes in Stage 3B Coats Disease
Table 3. Clinical Course of Patients with Stage 3B Coats Disease Treated with No Vitrectomy (>1 Year after Presentation) Total Presenting Initial Visual Follow-up Final Visual Case Age (mos) Acuity (mos) Acuity 11 13
39 42
CF CF 2’
120 72
CF 20/400
14
54
LP
60
LP
15
57
LP
108
NLP
Initial Ablation Treatment(s) (No.) SB/cryotherapy Photocoagulation, cryotherapy SB/cryotherapy/XD, photocoagulation (4) Cryotherapy (3), photocoagulation (1)
Post-ablation Visual Acuity TRD
Final Retinal Outcome
NVG Phthisis
CF CF
No No
Attached, macular scar Attached, macular scar
No No
No No
LP
No
Attached
No
No
LP
No
Subretinal hemorrhage and cholesterol crystals, pigmentary changes, retinal atrophy
No
No
CF ¼ counting fingers; LP ¼ light perception; NLP ¼ no light perception; NVG ¼ neovascular glaucoma; SB ¼ scleral buckling; TRD ¼ traction retinal detachment; XD ¼ external drainage of subretinal fluid.
development of this traction, which would provide a theoretic protective benefit in the later cicatricial phase of stage 3B Coats disease, when vitreoretinal traction begins to form. This is all balanced by the fact that there were 4 of 16 patients who did not require any vitrectomy because they did not develop any significant vitreoretinal traction at any point. Of note, 2 of these 4 patients did receive an SB early in their treatment regimen. It is not clear if this had any effect on the lack of need for a vitrectomy later. In this age of intravitreal injections, anti-VEGF agents and triamcinolone have emerged as adjuncts to the traditional methods of ablative treatment; anti-VEGF therapy has been shown to decrease macular edema and exudates5 and even reverse a total exudative detachment.6 However, the use of intravitreal bevacizumab and ranibizumab in Coats disease was thought to be possibly associated with the development of vitreoretinal fibrosis and TRD.7,8 This adverse effect has also been reported in the context of antiVEGF treatment of other ischemic retinopathies.8 In our series, 6 of 16 patients developed vitreoretinal fibrosis, of whom 5 developed TRDs. In only 1 of these patients was anti-VEGF treatment given, and it was given at the time of vitrectomy with the patient developing epiretinal membranes at the time of his last examination, seemingly unrelated to the anti-VEGF use. In the other 5 patients to whom no anti-VEGF was given, vitrectomy was performed when the TRDs became clinically significant at a mean of 2.6 years after onset of TRD, which in turn occurred a mean 20 months after initial ablative treatment of the ERD. According to the Shields classification, Coats disease is not characterized by vitreoretinal fibrosis, yet the late development of TRDs in the absence of anti-VEGF in these 5 patients suggests that this could represent a feature of advanced Coats disease. Vitreoretinal fibrosis and subsequent TRD have been reported earlier in the course of advanced Coats disease subsequent to intravitreal injection of bevacizumab8 and ranibizumab.7 Ramasubramanian and Shields8 reported 8 eyes with stage 2 to 3B Coats disease treated with intravitreal bevacizumab at the same time as standard treatment of laser photocoagulation, cryotherapy, and subTenon’s triamcinolone; 50% (4/8) developed vitreous
fibrosis at a mean of 5 months post-treatment, and 3 of the 4 progressed to TRD.8 Gaillard et al7 presented 5 cases of stage 3B Coats disease and 4 cases of stage 4 Coats disease treated with subretinal drainage to reform the vitreous cavity, with some patients also having intravitreal ranibizumab and subsequent treatment with photocoagulation and cryocoagulation. Overall, their mean follow-up was 46 months, with 5 of 9 developing vitreoretinal fibrosis, and 1 of whom developed a TRD followed by phthisis bulbi.7 In the subset of stage 3B Coats disease in the series by Gaillard et al,7 40% (2/5) developed vitreoretinal fibrosis at a mean of 15 months after ranibizumab injection. However, the mean follow-up for their subset of stage 3B patients was 40 months, with only 3 of these 5 followed for more than 3 years. In comparison, all 16 patients in our series had at least 5 years of follow-up. Of the 6 cases with vitreoretinal fibrosis in our series, 1 had antiVEGF use; however, the formation of epiretinal membranes in this patient was noted at the final follow-up 10 years later. Given this timing, it is unlikely that anti-VEGF played a role in the formation of epiretinal membranes in this case. The other 5 cases had progressive vitreous organization with eventual TRD after initial management Table 4. Timing of Onset of Vitreoretinal Fibrosis and Traction Retinal Detachment in Patients with Stage 3B Coats Disease Who Developed Vitreoretinal Fibrosis Time from Resolution Time from Onset of Active Lesions of TRD to Clinically Time from after Ablation to Significant TRD and Presentation to Case Onset of TRD (Mos) Vitrectomy (Mos) Vitrectomy (Mos) 2 16 3 4 5 1
4 N/A* 23 27 17 27
3 N/A 10 9 24 112
7 7 33 36 41 139
N/A ¼ not available; TRD ¼ traction retinal detachment. *Vitreoretinal fibrosis in the form of epiretinal membranes occurred late (10 years) after vitrectomy. See text for details.
5
Ophthalmology Volume -, Number -, Month 2017 without vitrectomy, with a mean 20 months from presentation to initial signs of vitreoretinal fibrosis and another mean 2.4 years for fibrosis to progress to a clinically significant TRD requiring vitrectomy. Subretinal and vitreoretinal fibrosis in Coats disease, whether in the macula or outside of the macula, is associated with poorer visual outcomes with the prevalence estimated to be 23%9 and 41%,10 respectively. Daruich et al10 reported 69 cases of Coats disease, ranging in severity from stage 1 to 4; their treatment protocol included laser photocoagulation targeting the peripheral telangiectasia, cryotherapy, subretinal fluid drainage, and intravitreal anti-VEGF in cases of severe ERD. They noted that 41% (28/69) of eyes with Coats disease developed extramacular fibrosis at a mean of 17 months after diagnosis (mean follow-up, 80 months). The onset of fibrosis noted by Daruich et al10 is close to the onset noted by Gaillard et al7 of 15 months and our finding of the onset of TRD at 20 months after ablative treatment. These instances of extramacular fibrosis can be further classified as vitreoretinal fibrosis alone (6/28), subretinal fibrosis alone (2/28), and vitreoretinal fibrosis with subretinal fibrosis (20/28). Among these 28 patients with extramacular fibrosis, 11 of 28 developed TRDs, which they found to be associated with severe vision loss; 63% (7/11) of eyes with TRD presented with no light perception at their last follow-up, compared with 3 of 58 who had no TRD had no light perception in their study.10 The pathogenesis of fibrosis in Coats disease has yet to be fully elucidated, with some suggesting that it could represent the natural course of the disease, given that they encountered vitreoretinal fibrosis before any treatment, antiVEGF or otherwise.11 In the study by Daruich et al,10 the number of patients receiving intravitreal bevacizumab or ranibizumab did not differ significantly between patients who developed extramacular subretinal or vitreoretinal fibrosis and those who did not. In our study, 5 of 6 patients developing vitreoretinal traction did not receive anti-VEGF therapy. In fact, overall anti-VEGF was used sparingly, ultimately in the late management of 2 patients. Its use in 1 case (case 6) was for the management of progression to stage 4 Coats disease to control anterior segment neovascularization and thus not relevant to the discussion of pathogenesis of vitreoretinal traction. In the other case (case 16), the patient received anti-VEGF treatment twice and developed epiretinal membranes at the end of follow-up. Given the incidence of vitreoretinal fibrosis in the absence of anti-VEGF use in 5 cases in this series, this supports the idea that vitreoretinal fibrosis and tractional detachment represents the end-stage in the natural history of exudative Coats disease as posited by Zheng and Jiang,11 as opposed to an adverse effect of anti-VEGF treatment as proposed by Ramasubramanian and Shields.8 In analyzing the various factors that contribute to extramacular subretinal or vitreoretinal fibrosis, Daruich et al10 also showed that the type of ERD, whether total, partial, or none, and the degree of retinal exudation, measured by clock-hours, were significant independent contributing factors to the development of extramacular subretinal or vitreoretinal fibrosis. Furthermore, the extent of retinal exudation, a defining characteristic of Coats disease, and
6
degree of ERD were shown to be significantly higher in eyes with extramacular subretinal or vitreoretinal fibrosis.10 It is evident from the literature that both subretinal fibrosis and vitreoretinal fibrosis are linked to poor outcomes. A few case reports have measured a high level of VEGF in the vitreous, subretinal fluid, and aqueous of eyes with Coats disease, which subsequently decreased after intravitreal anti-VEGF treatment along with some clinical improvement.12,13 These reports suggest a role for the dysregulation of VEGF in the pathogenesis of Coats disease and a rationale for anti-VEGF agents in the treatment of Coats disease. Nevertheless, on the basis of the experience of Ramasubramanian and Shields,8 regimens involving antiVEGF in Coats disease may need to be optimized with the other modalities to maximize vision and globe salvage. In terms of timing of vitrectomy with respect to the incidence of vitreoretinal traction, 2 of 8 patients who underwent early PPV (within the first year of presentation) developed vitreoretinal traction, whereas 4 of 8 in the expectant management (late PPV or no PPV) had developed vitreoretinal traction. In the 2 cases that underwent early PPV, 1 was for the treatment of TRD that developed at 7 months after presentation, whereas the other had EV and developed epiretinal membranes in a delayed fashion approximately a decade after PPV. The other 6 patients who underwent early PPV did not show signs of subsequent vitreoretinal fibrosis. In the group of patients in whom PPV was delayed until judged to be necessary, 4 of 8 required PPV for progressive vitreoretinal traction that developed into a late-onset clinically significant TRD. Although the numbers were small, we did see that there were fewer cases of vitreoretinal traction in the group that underwent early (prophylactic) vitrectomy. However, a conclusion of benefit cannot be firmly reached because of the nature of this retrospective study and has to be balanced by the fact that 4 of 8 of the expectant management group never required any vitrectomy for vitreoretinal traction. As expected, most of the patients in our series were male. However, the occurrence of advanced Coats disease in a girl with Turner syndrome (case 4) merits further consideration. Although Coats disease is not a common finding in Turner syndrome, cases of patients with both conditions have been reported in the literature.14e16 Once thought to be a “chance occurrence of the 2 syndromes,”14 the presence of Coats disease in a patient with Turner syndrome could represent another vascular anomaly associated with Turner syndrome, such as aortic coarctation, cerebral vessel aneurysm, and gastrointestinal telangiectasia.15 It may be that additional investigation into the pathogenesis of these vascular anomalies in Turner syndrome could advance our understanding of Coats disease. The limitations of this study include its retrospective nature and that treatment was not standardized across surgeons. Nevertheless, it is significant in that it presents longterm follow-up in stage 3B Coats disease. Overall, the globe was salvaged in 15 of 16 cases with a mean follow-up of 9.5 years. These interventions with long-term follow-up could not reverse the poor visual outcomes associated with advanced Coats disease. These treated eyes do better than the natural history of stage 3B Coats disease. In the majority
Li et al
Long-Term Outcomes in Stage 3B Coats Disease
of cases, treatment was successful in preserving the globe without pain. In the patient in whom enucleation became necessary, these interventions allowed the globe to remain for several more years than otherwise expected without treatment and allowed the orbit to mature further to better accommodate a prosthesis in the future. Vitrectomy at the time of initial treatment may be beneficial in preventing delayed onset vitreous organization and TRD, although a larger prospective study comparing treatments looking at the development of fibrosis in Coats disease is necessary to direct management and possible timing of vitrectomy. Furthermore, a better understanding of fibrosis in the progression of advanced Coats disease would be helpful in guiding management and possibly developing improved interventions.
References 1. Shields JA, Shields CL, Honavar SG, et al. Classification and management of Coats disease: the 2000 Proctor Lecture. Am J Ophthalmol. 2001;131:572e583. 2. Shields JA, Shields CL, Honavar SG, Demirci H. Clinical variations and complications of Coats disease in 150 cases: the 2000 Sanford Gifford Memorial Lecture. Am J Ophthalmol. 2001;131:561e571. 3. Ghorbanian S, Jaulim A, Chatziralli IP. Diagnosis and treatment of Coats’ disease: a review of the literature. Ophthalmologica. 2012;227:175e182. 4. Silodor SW, Augsburger JJ, Shields JA, Tasman W. Natural history and management of advanced Coats’ disease. Ophthalmic Surg. 1988;19:89e93. 5. Sigler EJ, Randolph JC, Calzada JI, et al. Current management of Coats disease. Surv Ophthalmol. 2014;59:30e46.
6. Zhao T, Wang K, Ma Y, Jiang YR. Resolution of total retinal detachment in Coats’ disease with intravitreal injection of bevacizumab. Graefes Arch Clin Exp Ophthalmol. 2011;249: 1745e1746. 7. Gaillard MC, Mataftsi A, Balmer A, et al. Ranibizumab in the management of advanced Coats disease Stages 3B and 4: longterm outcomes. Retina. 2014;34:2275e2281. 8. Ramasubramanian A, Shields CL. Bevacizumab for Coats’ disease with exudative retinal detachment and risk of vitreoretinal traction. Br J Ophthalmol. 2012;96: 356e359. 9. Jumper JM, Pomerleau D, McDonald HR, et al. Macular fibrosis in Coats disease. Retina. 2010;30:S9e14. 10. Daruich A, Matet A, Tran HV, et al. Extramacular fibrosis in Coats’ disease. Retina. 2016;36:2022e2028. 11. Zheng XX, Jiang YR. The effect of intravitreal bevacizumab injection as the initial treatment for Coats’ disease. Graefes Arch Clin Exp Ophthalmol. 2014;252:35e42. 12. He YG, Wang H, Zhao B, et al. Elevated vascular endothelial growth factor level in Coats’ disease and possible therapeutic role of bevacizumab. Graefes Arch Clin Exp Ophthalmol. 2010;248:1519e1521. 13. Sun Y, Jain A, Moshfeghi DM. Elevated vascular endothelial growth factor levels in Coats disease: rapid response to pegaptanib sodium. Graefes Arch Clin Exp Ophthalmol. 2007;245:1387e1388. 14. Cameron JD, Yanoff M, Frayer WC. Coats’ disease and Turner’s syndrome. Am J Ophthalmol. 1974;78:852e854. 15. Beby F, Roche O, Burillon C, Denis P. Coats’ disease and bilateral cataract in a child with Turner syndrome: a case report. Graefes Arch Clin Exp Ophthalmol. 2005;243: 1291e1293. 16. Desai RU, Saffra NA, Krishna RP, Rosenberg SE. Coats’ disease, Turner syndrome, and von Willebrand disease in a patient with Wildtype Norrie disease pseudoglioma. J Pediatr Ophthalmol Strabismus. 2011;48 Online:e1e3.
Footnotes and Financial Disclosures Originally received: May 18, 2017. Final revision: November 30, 2017. Accepted: December 5, 2017. Available online: ---. 1 Hofstra Northwell School of Medicine, Great Neck, New York. 2
Associated Retinal Consultants, Royal Oak, Michigan.
3
Cleveland Clinic, Cleveland, Ohio.
4
KYDOFT Foundation, Santiago, Chile.
5
Long Island Vitreoretinal Consultants, Great Neck, New York.
Presented as a poster at: the American Academy of Ophthalmology Annual Meeting, November 11e14, 2017, New Orleans, Louisiana. Financial Disclosure(s): The author(s) have made the following disclosure(s): P.J.F.: Consultant Genentech, Allergan, Alcon, and Santeen. HUMAN SUBJECTS: This study includes human subject/tissues. Study protocol was approved by IRB of North Shore University Hospital. All tenets of the Declaration of Helsinki were followed.
Author Contributions: Conception and design: Li, Ferrone Data collection: Li, Capone Jr., Trese, Sears, Kychenthal, De la Huerta, Ferrone Analysis and interpretation: Li, Ferrone Obtained funding: Not applicable Overall responsibility: Li, Capone Jr., Trese, Sears, Kychenthal, De la Huerta, Ferrone Abbreviations and Acronyms: ERD ¼ exudative retinal detachment; EV ¼ early vitrectomy; NVG ¼ neovascular glaucoma; PPV ¼ pars plana vitrectomy; TRD ¼ traction retinal detachment; VEGF ¼ vascular endothelial growth factor; XD ¼ external drainage of subretinal fluid. Correspondence: Philip J. Ferrone, MD, Long Island Vitreoretinal Consultants, 600 Northern Boulevard, Suite 216, Great Neck, NY 11021. E-mail: PJFerrone@gmail. com.
7