Intravitreal Bevacizumab for Choroidal Neovascularization Associated with Angioid Streaks: Long-term Results.

PURPOSE: The purpose of the study was to evaluate the long-term efficacy and safety of intravitreal bevacizumab (IVB) in the treatment of choroidal neovascularization (CNV) secondary to angioid streaks (AS).
MATERIALS AND METHODS: Noncomparative, interventional retrospective case series involving ten patients (15 eyes) with a minimum follow-up of 25 months following IVB for AS-associated CNV. Demographic and clinical details at baseline and during follow-up were collected from patient records. Detailed clinical examination was followed by fundus fluorescein angiography and optical coherence tomography to confirm CNV. Both primary and recurrent CNVs were treated with monthly injections of IVB till the lesion stabilized. Primary outcome measures were the percentage of patients with stable or improved visual acuity (VA) and with stable or decreased central retinal thickness (CRT) at the last visit. Secondary outcome measure was the incidence of ocular and systemic complications.
RESULTS: Recurrence was observed in 11 eyes (73.33%) over a mean follow-up of 57.33 months (range: 25-100). A mean number of injections administered was 5.60. VA improved or stabilized in 73.33% and deteriorated in 26.67% of eyes at the final visit. Mean CRT improved from 324.40 μm at baseline to 265.53 μm at final visit, which was statistically significant. Complications observed were ocular hypertension in one patient and thromboembolic event in another patient.
CONCLUSIONS: IVB appears to be a safe and effective option to treat CNV and to preserve vision over a prolonged period. It cannot eliminate the risk of recurrent CNV indicating the need for more effective treatments to arrest this visually debilitating condition.

Introduction

Angioid streaks (ASs) are reddish brown, jagged lines radiating from peripapillary area into the peripheral fundus, predominantly affecting population in the second to fifth decade. Systemic associations include pseudoxanthoma elasticum (PXE), Ehler-Danlos syndrome, Paget's disease, or hemoglobinopathies.[12] Choroidal neovascularization (CNV) complicating AS (prevalence: 72%–86%) is an important cause of legal blindness and financial burden in middle-aged individuals affected by this condition.[3] Treatment of CNV has remained a challenge, and different modalities such as laser photocoagulation, transpupillary therapy, photodynamic therapy (PDT), or surgery usually resulted in poor outcomes.[145]

Vascular endothelial growth factors (VEGFs), a potent mitogen for endothelial cells, have been implicated in the pathogenesis of CNV.[4] A recent review on the treatment of AS-related CNV has shown Anti-VEGF drugs such as bevacizumab and ranibizumab to be the most effective treatment.[1] Anti-VEGF therapy in angioid CNV was first reported in 2006.[6] Subsequently, several case reports and case series have confirmed its efficacy and safety over a period ranging from 6 to 24 months.[47891011121314] A recent prospective multicentric study involving 44 patients with AS-associated CNV also confirmed the efficacy of Anti-VEGF therapy in stabilizing CNV over a period of 12 months.[5] However, AS being a chronic disease affecting young individuals, it is important to understand the long-term benefits of anti-VEGF therapy, and so a retrospective analysis of our ten patients (15 eyes) treated with intravitreal bevacizumab (IVB) and with a minimum follow-up of 25 months (range: 25–100 and mean: 57.33 months) was done. To the best of our knowledge, this is the longest follow-up results reported for this rare condition.

Materials and Methods

This is a retrospective chart review of ten patients (15 eyes) treated for CNV secondary to AS in the Retina Services of our hospital from September 2006 to March 2014 and who had completed a minimum follow-up of 25 months. Data collected from the patient's chart included age, gender, details of clinical evaluation, and treatment at baseline and subsequent visits. A total number of injections given, duration of total follow-up (from the initial treatment visit to the last visit), and duration of follow up after the last intervention and any local or systemic adverse effects of treatment were also noted. Ophthalmic evaluation included checking of best-corrected visual acuity (VA) with Snellen's chart, slit lamp biomicroscopy, applanation tonometry, indirect ophthalmoscopy followed by fundus fluorescein angiography (FFA) and optical coherence tomography (OCT). FFA was performed using Zeiss FF450; Zeiss Oberkochen, Germany, and OCT was done using time domain Stratus OCT or spectral domain Cirrus HD-OCT (Carl Zeiss Meditech, USA). Patients were also evaluated by general physician to look for systemic associations of AS. CNV identified at the time of presentation was labelled as primary CNV and its location and activity were evaluated based on the leakage seen on FFA and the morphological changes on OCT such as presence of CNV, subretinal fluid (SRF), and retinal edema. All patients with active CNV and with no systemic morbidities such as uncontrolled hypertension, cardiovascular, or cerebrovascular events in the previous 6 months were administered IVB 1.25 mg in 0.05 ml (Avastin; Genentech, South San Francisco, CA, USA) under aseptic precautions, as an outpatient procedure. Antibiotic eye drops four times a day for 5 days were prescribed after the injection for all patients. The study adhered to the tenets of the Declaration of Helsinki and was approved by the Institutional Ethics Committee. An informed written consent was obtained from all patients.

Patients were followed up at monthly intervals with reinjections on as needed basis till CNV became inactive. At each visit, recording of VA, fundus examination, fundus photography, and OCT was done. FFA was repeated only when a new lesion was suspected or in eyes with poor response to therapy. Improvement or stabilization of VA along with resolution of CNV, retinal edema, and SRF on OCT was considered as a favorable response to therapy. Recurrent CNV was diagnosed if deterioration of vision and/or metamorphopsia was associated with the presence of new hemorrhage or active lesion adjacent to primary CNV or at new foci with the corresponding leakage on FFA and reappearance of SRF or retinal edema on OCT. Additional injections were administered to treat recurrent CNV and monitoring continued at monthly intervals till stabilization was achieved. This was in accordance with the retreatment criteria for neovascular age-related macular degeneration which allows less frequent injections with comparable outcome.[15] Two patients also underwent PDT. This included the first patient in this series who presented with a subfoveal CNV in his left eye (OS) in September 2006, when the standard of care for subfoveal CNV was PDT. IVB was a new option, and its efficacy as monotherapy was not well established. Hence, he was initially treated with a combination of PDT and IVB, but the multiple recurrences in that eye as well as the CNV that developed later in the fellow eye were managed with IVB alone. Similarly, patient 5 initially managed by IVB monotherapy for CNV in the OS, preferred PDT when he was detected later to have bilateral active CNV, to reduce the need for multiple injections in both eyes. Following treatment, OS stabilized with a large subfoveal scar, whereas right eye (OD) developed three more recurrences which were managed by IVB monotherapy.

Primary end points of this study were the percentage of patients with stable or improved VA and with stable or decreased macular thickness at the last visit. Secondary end point was the long-term incidence of ocular and systemic complications in patients with AS who are predisposed to systemic complications. For statistical analysis, Snellen's VAs were converted to logMAR. Any change in VA was defined as improved or worse based on gain or loss of 3 or more lines from baseline VA. If change in vision was within three lines of baseline VA, it was considered as stable. Morphological outcome was analyzed by comparing the mean central retinal thickness (CRT) measured on central macular map on OCT at baseline and at final visit. Data were entered in Excel and analyzed using IBM SPSS software version 19.0. Continuous variables were expressed in mean and standard deviation. Quantitative data were compared using Paired “t” test, and correlation between baseline and final VA and CRT was done using Pearson correlation. P < 0.05 was considered statistically significant.

Results

Fifteen eyes of ten patients treated for AS-associated CNV and with a minimum follow-up of 25 months were analyzed [Table 1]. There were seven males and three females with a mean age of 49.47 ± 4.37 years (range: 46–63 years). Three patients had biopsy-proven PXE, and in others, AS was considered as idiopathic. Follow-up period ranged from 25 to 100 months with a mean of 57.33 ± 21.76 months. Mean follow-up after the last injection was 28.26 ± 23.42 months (range: 0–73). CNV was bilateral in seven patients (70%), of which five had active CNV bilaterally, and two had active CNV in one eye and scarred CNV in the fellow eye. All the three patients with PXE had bilateral CNV, but in two of them, CNV was already scarred in one eye and was not included in the study. All eyes had naïve CNV, of which eight (53.33%) were juxtafoveal and 7 (46.67%) were subfoveal CNV.

Table 1

Demographic data, treatment details and visual and morphological data at baseline and final visit

A mean number of injections needed to treat primary CNV was 2.26 ± 0.96. Three eyes (20%) stabilized with one injection, seven eyes (46.67%) with two injections, three eyes (20%) with three injections, and two eyes (13.33%) with four injections. Thus, primary CNV resolved with ≤3 injections in 83.67% of eyes. Mean baseline CRT was 283.33 ± 17.21 μm in eyes that responded to one injection compared to 473.50 ± 106.78 μm in eyes which needed four injections. Recurrence was observed in 11 eyes (73.33%) whereas 4 eyes (26.67%) remained inactive throughout follow-up. Most of the recurrences developed adjacent to the primary CNV. Time of detection of the first recurrence ranged from 10 to 26 months. Six eyes had only one recurrence, two eyes had two recurrences, one eye had three recurrences, and two eyes had four recurrences. Number of injections required to obliterate recurrent CNV were 11 injections (in 1 eye), 10 injections (in 1 eye), 5 injections (in 4 eyes), 2 injections (in 4 eyes), and 1injection (in 1 eye) amounting to a total of fifty injections (mean 3.33 ± 3.50). Overall, the mean number of injections administered was 5.60 ± 3.50. There was a statistically significant correlation between the mean number of recurrences to the number of injections given for recurrent CNV (P = 0.00) as well as to the total number of injections (P = 0.00). All the three eyes treated with PDT had poor outcome, with two of these eyes having the maximum recurrences.

VA improved or stabilized in 73.33% of eyes (11 eyes) and deteriorated in 26.67% of eyes (four eyes). However, the mean logMAR VA worsened from 0.68 ± 0.54 at baseline to 0.71 ± 0.52 at final visit, but the difference was not statistically significant (P = 0.75). Mean variation in VA was −0.27 ± 3.63 lines ranging from −6 lines to +9 lines. At baseline, 26.67% (four eyes) had VA 20/40 (logMAR: 0.3) or better which improved to 33.34% (five eyes) at final visit. Similarly, proportion of eyes with VA 20/200 (logMAR: 1.0) or worse also increased from 33.34% at baseline to 40% at final visit. However, no statistically significant conclusions could be derived as the numbers were small. Mean CRT improved from 324.40 ± 88 μm at baseline to 265.53 ± 66.41 μm at the final visit, a difference which approached statistical significance (P = 0.01). Following treatment, 14/15 eyes (93.33%) had stable or decreased CRT with inactive CNV. The only exception was the OD of patient 5 (treated with PDT and IVB) which developed multiple recurrences and chronic cystoid macular edema (CME) with worsening of CRT from 289 μm to 409 μm. Serial morphological changes in patients with the best and worst outcomes are shown in Figures 1 and 2, respectively. Overall, ten eyes developed macular scar which was associated with CME in three eyes, foveal thinning in two eyes, and vitreomacular traction in one eye. Thus, despite an improvement in the retinal thickness, majority had poor functional outcome due to degenerative macular changes.

Figure 1

Images of patient 2 who had successful outcome following intravitreal bevacizumab for choroidal neovascularization in the right eye. At presentation, color photograph (CF) (a) and fundus fluorescein angiography (b) showed angioid streak (white arrow) and Juxtafoveal choroidal neovascularization with hemorrhage (black arrow). Optical coherence tomography (c) showed choroidal neovascularization with subretinal fluid, which resolved after treatment as seen on CF (d), fundus fluorescein angiography (e), and optical coherence tomography (f). 10 months later, recurrent choroidal neovascularization was detected based on fresh hemorrhage on CF (g), mottled hyper fluorescence on fundus fluorescein angiography (h), and subretinal fluid on optical coherence tomography (i) and was treated with one more injection. He remained stable at his final visit with a scarred juxtafoveal choroidal neovascularization clinically (j) and on optical coherence tomography (k)

Figure 2

Images of patient five who had the worst outcome. At presentation, color photograph (CF) (a and b), fundus fluorescein angiography (c and d) showed angioid streak (white arrow) in both eyes and choroidal neovascularization (black arrow) in left eye. Optical coherence tomography (e and f) of right eye was normal, left eye had choroidal neovascularization which resolved after treatment (g). 20 months later (h and i), there was new choroidal neovascularization in right eye and reactivation in left eye which stabilized with treatment (j and k). Multiple recurrences in right eye caused chronic edema and scarring as observed in the pre (l and n) and post (m and o) injection scans. CF (p and q), fundus fluorescein angiography (r and s) and optical coherence tomography (t and u) at his final visit showed partially scarred choroidal neovascularization with active inferotemporal edge in right eye and fibrotic scar in left eye

The only ocular complication noted during follow-up was increased intraocular pressure in one eye (patient 1, OS) which was controlled with Timolol eye drops. Patient 3, who did not have PXE but a known diabetic, hypertensive, and cardiac patient under our care since 2008, developed minor cerebrovascular accident in April 2012 and also had to undergo coronary artery bypass graft in February 2014. No other ocular or systemic complications were observed during the follow-up.

Discussion

All the published studies regarding IVB in AS-associated CNV and having a minimum follow-up of 23 months are summarized in Table 2. Alagöz et al.[16] who have the largest series achieved improvement or stabilization of VA in 65.2% of their eyes, but 61% (14/23 eyes) continued to have active CNV at the last follow-up. Young patients with resistant CNV and those with PXE required more aggressive treatment. However, there are reports with a lesser number of patients that have demonstrated 100% improvement/stabilization both visually and morphologically over a follow-up period ranging from 23 to 28.6 months.[101718] Finger et al.,[19] in their analysis of 14 patients with PXE and AS, found visual improvement in those with early disease and emphasizes the need for early initiation of treatment with regular follow-up and prompt reinjection for recurrence. Maintenance therapy with monthly injections is not necessary considering the cardiovascular risk profile of these patients with PXE and also to minimize the risk of RPE atrophy which can be aggravated by repeated anti-VEGF injections. Myung et al.[18] treated nine patients with PXE and AS-related CNV and reported statistically significant improvement in the final VA at 6 months in their preliminary study. However, in their extended study of the same cohort of patients followed up for 28.6 months, overall visual outcome was less favorable due to frequent recurrences and scarring. Similar observations were also made by Iacono et al.,[20] who followed up 15 patients with nonsubfoveal CNV secondary to AS and found that the efficacy of IVB in halting the CNV deteriorated from 80% at 12 months to 46% at 36 months. Another recent study by Rosina et al.[21] evaluating 16 eyes with a mean follow-up of 52 months states that risk of recurrence increased during the first 50 weeks and remains stable afterward. More injections were needed in eyes with increased CNV reactivation, and the final VA is related to the initial VA and not to the number of injections, previous treatments, or baseline CNV size. In a comprehensive review of treatment of AS-related CNV, Gliem et al.[1] also support that potential benefit of anti-VEGF therapy correlates best with the structural alterations and functional loss present at baseline. Based on a pooled analysis of 16 studies on the efficacy of bevacizumab or ranibizumab, authors conclude that there is no superiority of one drug over the other and also of fixed versus as needed regimen.

Table 2

Published data regarding long-term efficacy of intravitreal bevacizumab in choroidal neovascularization secondary to angioid streaks

We preferred bevacizumab in our patients due to its affordability and proven efficacy in CNV. In our series with the longest follow-up reported till date (57.33 ± 21.76 months), there was stabilization or improvement in VA in 73.33% and morphological stabilization in 93.33% of patients. These results were achieved with a mean of 5.60 ± 3.50 injections. Our results are comparable to the long-term outcomes with IVB reported by other authors and far superior to the outcome achieved with the treatments in the pre- anti-VEGF era.[1] Recurrent CNV was observed in 73.33% and required more treatment for stabilization (mean of 4.54 injections) than for the primary CNV (mean of 2.26 injections). First recurrence was noted at 10 months and the last recurrence at 87 months in the patient with 100-month follow-up [Figure 3]. Interval between recurrences ranged from 5 months (Patient 5, OD) to 32 months (Patient 1, OS and Patient 2, OD) indicating that reactivation of the CNV can continue to occur at any time and at any location in some eyes in this chronically active disease necessitating long-term monitoring with timely intervention. Here, we differ from Rosina et al.,[21] who observed that patients free from disease reactivation showed a smaller recurrence rate. Furthermore, the same authors did not find any difference between those with naïve CNV and those treated with PDT. However, all the three eyes treated with PDT in our study had a recurrence with worse vision at final visit. Abdelkader et al. also have reported disappointing results with PDT and recommend anti-VEGF therapy as the treatment of choice at the present time for CNV complicating AS.[5] Since only three patients in our series had PXE, we could not compare the difference in the outcome between those with and without PXE. The reported ocular complications with bevacizumab include lens touch by the needle[12], Bruch's membrane rupture,[22] and arteritic anterior ischemic optic neuropathy.[23] In our series, ocular hypertension was detected in one patient which was well controlled with a single antiglaucoma medication and never developed any progressive glaucomatous optic neuropathy. Besozzi et al.[24] have reported acute stoke within 3 days of injection in a patient with PXE and no other systemic illness. Although one patient in our cohort developed two thromboembolic events, it is difficult to establish a causal relationship with IVB as he had other risk factors such as systemic hypertension and diabetes mellitus. Majority of the authors did not observe any ocular or systemic complications.[3151719]

Figure 3

Images of the left eye of patient 1 which had the longest follow-up. At presentation, color photograph (CF) (a) and fundus fluorescein angiography (b) showed angioid streak (white arrow) and juxtafoveal choroidal neovascularization (black arrow). Optical coherence tomography (c) showed choroidal neovascularization with subretinal fluid. Posttreatment Optical coherence tomography (d and e) showed gradual resolution of subretinal fluid (yellow arrow). Serial fundus photographs with optical coherence tomography images pre- and post-treatment of his first recurrence (f-h), second (i-k), third (l-n), and fourth (o-q) demonstrate the progressive macular scarring. He remained stable at his last visit with a scarred choroidal neovascularization clinically (r) and on fundus fluorescein angiography (s). Thinning of neurosensory retina was seen on optical coherence tomography (t)

The strength of our study is the long duration of follow-up involving a significant number of eyes which is the longest reported so far, for this relatively rare condition. The limitations include retrospective design and heterogeneous study population. Despite these limitations, our results confirm the long-term efficacy of anti-VEGF therapy to stabilize the CNV and to preserve or even improve vision. It also confirmed the long-term safety of bevacizumab in patients with AS. However, it cannot eliminate the risk of recurrent CNV indicating the need to develop more effective strategies to arrest this visually debilitating condition.

Conclusions

IVB is a safe and effective option in the long-term stabilization of CNV due to AS.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.