Optic Coherence Tomography Angiography Findings of Bilateral Choroidal Neovascularization Associated with Optic Disc Drusen Treated with Intravitreal Aflibercept Injection

Purpose The purpose of this case report is to present the findings of optical coherence tomography angiography (OCTA) of a patient with bilateral choroidal neovascularization (CNV) associated with optic disc drusen (ODD), who was treated with intravitreal aflibercept injection. Case presentation. A 14-year-old girl presented with a complaint of visual loss and metamorphopsia in her both eyes. Best-corrected visual acuity (BCVA) was 20/32 and 20/25, respectively, in the right and left eyes. Intraocular pressure and anterior segment examination were normal. Dilated fundus examination revealed elevated optic discs with blurred margins in both eyes. In addition, slightly elevated yellow lesion extending from optic nerve head to the macula was observed bilaterally. The patient underwent imaging with colour fundus photography, fundus autofluorescence (FAF), fundus fluorescein angiography (FFA), spectral-domain optic coherence tomography (SD-OCT), OCTA, orbital ultrasonography (USG), and computed tomography (CT). In particular, OCTA demonstrated clearly the large circular CNV complex in the right eye and the CNV structure in the left eye containing slightly activated main trunk and minimal vessel loops in the papillomacular region. CNV secondary to bilateral ODD was suspected. Intravitreal aflibercept injections were performed in 3 doses to the right eye and a single dose to the left eye. After the injections, BCVA reached its complete level in both eyes. SD-OCT revealed irregularity of RPE in the temporal region of the optic disc and complete regression of the subretinal fluid. Interestingly, the entire CNV complex including the main trunk completely disappeared in OCTA. CNV complex was not observed in OCTA during 1-year follow-up, and peripapillary and macular vascular density measurements did not show any significant change. BCVA was preserved, and no additional injections were needed. Conclusion It is possible that OCTA can be used for detailed evaluation of CNV associated with ODD, response to anti-VEGF treatment, and peripapillary and macular vascular density. There is a need for further studies to confirm the changes such as disappearance of CNV in OCTA after injection as we observed in our patient.


Introduction
Optic disc drusen (ODD) are calcified hyaline-like deposits in the optic nerve head that are mainly located in front of the lamina cribrosa. They are often bilateral (67%-91%), and their incidence is 0.4%-20.4% in general population while there is female preponderance in their prevalence. They are usually benign and vision sparing; however, they may rarely complicate with visual field defects, haemorrhages, choroidal neovascularization (CNV), serous maculopathy, vascular occlusion, and nonarteritic ischemic optic neuropathy (NAION) [1].
Noninvasive imaging of the retina, choroidal, and disc microvasculature is now possible with OCTA and without the use of exogenous intravenous dye injection. It can provide information for the evaluation of both structure and blood flow. It particularly allows the assessment of CNV morphology and features activity as well as the timing of and response to anti-VEGF treatment of CNV [3,4]. It has been increasingly used for the diagnosis and imaging of paediatric retinal vascular disease and especially CNV as it is both reliable and noninvasive [5].
The OCTA studies in the literature mainly focus on microvascular changes associated with optic disc drusen. To the best of our knowledge, only Ong et al. demonstrated inactivate CNV associated with ODD in 1 case with OCTA in a series of 8 paediatric patients with CNV. However, this is the first time to report an interesting change in CNV after injection and long-term follow-up findings.

Case Report
A 14-year-old girl was referred to us with complaints of visual loss and metamorphopsia in both eyes that had persisted for 1 week. Her best-corrected visual acuity (BCVA) was 20/32 in the right eye and 20/25 in the left eye. Her ocular and systemic history was unremarkable. Her intraocular pressure and anterior segment examination were normal. Dilated fundus examination revealed elevated optic discs with blurred margins in both eyes. In addition, slightly elevated yellow lesion extending from optic nerve head to the macula was observed bilaterally (Figures 1(a) and 1(d)). The patient underwent imaging with colour fundus photography, FAF, FFA, spectral-domain optic coherence tomography (SD-OCT, Cirrus HD OCT 5000, Carl Zeiss Meditec Inc., Dublin, CA, USA), OCTA, orbital USG, and computed tomography (CT). FAF showed bilateral aspect of white refractile and hyperautofluorescence bodies on the surface of the optic nerve linked with the drusen and a hypoautofluorescent and hyperautofluorescent lesion in the papillomacular region (Figures 1(b) and 1(e)). FFA showed no hyperfluorescence or leakage with blurred borders of the optic disc, and early hyperfluorescence and intense leakage that might be associated with CNV were observed in both eyes with the right eye predominance (Figures 1(c) and 1(f )). The peripheral retina was normal. B-mode USG showed hyperechogenic appearance, and orbital CT revealed papilla as bright spots which might be associated with optic disc drusen (Figure 1(g)). SD-OCT showed retinal thickening, subretinal hyperreflectivity in the papillomacular region bilaterally, and subfoveal fluid in the right eye (Figures 1(h) and 1(i)). Furthermore, macular and retinal nerve fiber layer (RNFL) thickness measurements were recorded (Table 1).
OCTA revealed superficial and deep capillary density, foveal avascular zone, flow, and en face images in the macula ( Table 2). A large circular CNV complex with main trunk, multiple dense thin capillaries branching from the main trunk in a tree-like manner, and frequent anastomoses was observed in the papillomacular region of the right eye in the outer retina and choriocapillaris cross-sections (Figures 2(a) and 2(b)). In the left eye, CNV with main trunk, minimal vessel loops, and capillaries was observed, which was considered as slightly activated (Figures 2(c) and 2(d)). Moreover, radial papillary capillary density was measured with OCTA ( Table 2). Density was lower in the areas associated with nasal quadrant compared to the other quadrants.
All evaluations suggested CNV secondary to ODD, and intravitreal aflibercept was injected in 3 doses to the right eye and a single dose to the left eye. Three doses were injected to the right eye every other month. After the first injection, BCVA remained as 20/32 and increased to 20/25 after the second one. After the third dose of injection to the right eye and the first one to the left eye, BCVA was 20/20 in both eyes. After the injections in both eyes, SD-OCT revealed RPE irregularity in the temporal region of the optic disc and complete regression of the subretinal fluid. OCTA interestingly showed that the entire CNV compete including the main trunk disappeared (Figures 3(a) to 3(d)). CNV complex was not observed with OCTA, and no significant change was observed in the peripapillary and macular vascular density measurements, and findings which might be correlated with RNFL measurements were not detected in 1-year follow-up. BCVA was preserved, and no additional injections were needed.

Discussion
CNV associated with ODD is a very rare complication, which has been reported in both adults and children [6]. In most of the patients, CNV membranes associated with ODD occur nasally; however, they may also occur temporally resulting in serous haemorrhagic maculopathy, cystoid macular edema, and macular scarring [7]. Its pathogenesis is unknown. Possible mechanisms involved in this complication include the compressive effect of drusen on the surrounding blood vessels, which leads to mechanical impairment of peripapillary vascular integrity, vascular congestion, or ischemia. Retinal ischemia and the release of VEGF may be the factors that trigger the development of CNV [8,9]. Treatment must be provided if vision is jeopardized. The modern antivascular endothelial growth factor (VEGF) medications, photodynamic therapy, and surgical removal have each been demonstrated to be successful [10]. Anti-VEGF agents have been increasingly preferred as the first-line treatment because they ensure high visual acuity and disease stabilization usually with fewer intravitreal injections especially in paediatric cases, and recurrence is almost never observed. Gan and Long reported complete resolution after 3 doses of injection in cases for whom aflibercept was injected for the first time [11]. Knape et al. reported rapid and atraumatic resolution with the combination of bevacizumab and focal laser photocoagulation [12]. Alkın et al. emphasized that CNV secondary to ODD might be more sensitive than CNV secondary to age-related macular degeneration (AMD) in cases injected with a single dose of ranibizumab [13]. We achieved rapid response and did not observe recurrence in cases injected with 3 doses of aflibercept in the right eye and a single dose 2 Case Reports in Ophthalmological Medicine of aflibercept in the left eye and whom we followed for 12 months with OCTA. 4 different CNV patterns associated with exudative agerelated macular degeneration can be detected with OCTA: Medusa pattern: central feeder vessel, circular peripheral anastomosis, thin branches, and hypointense halo; seafan pattern: eccentric feeder vessel, thin branches, and hypointense halo; indistinct network pattern: thin branches and hypointense halo; and pruned vascular tree pattern: persis-tence of main vascular trunks [14]. Pruned vascular tree pattern is considered as inactivate and usually followed up. The morphology of CNV and structural response to anti-VEGF treatment are strongly correlated. Anti-VEGF is ineffective for the vascular trunk while it is effective for especially the newly formed capillaries from preexisting vascular trunks while there is a correlation between better functional outcomes an increased vascularity of the CNV [15]. Spaide demonstrates reprofileration 20-50 days after the injection [16].

Case Reports in Ophthalmological Medicine
However, long-term inactivation can be ensured with fewer anti-VEGF injections in cases of CNV associated with high myopia, punctate inner choroidopathy (PIC), choroidal osteoma, optic nerve drusen like our patient, and idiopathic with lower activity than AMD [17][18][19][20]. The literature does not contain wide range of data which can be used to assess why stabilization can be achieved with fewer injections and make comparisons with CNV patterns of AMD with OCTA as highlighted above and assess reproliferation.
The most striking finding after injection in our case was the absence of CNV visualization in OCTA. The absence of CNV visualization in OCTA was associated with clinical and SD-OCT exudative inactivity [14]. In fact, inactive lesions are visualized as pruned vascular tree pattern, presence of large linear vessel with no or minimal anastomosis in OCTA. This may be due to other reasons such as the masking effect generated by a high PED, media opacities, or the presence of haemorrhages [21]. However, they were not observed in the follow-up of our patient. Hua and Ning demonstrated the development of atrophy in the CNV complex after 3 doses of anti-VEGF injections in PIC cases with high myopia and round dark region visualization due to no-perfusion/hypopefusion zone in the choriocapillaris [22]. This situation which was also observed in our case can be interpreted as decreased signal due to inactivity, but the absence of pruned vascular tree pattern may at least contribute to explaining why atrophy/shrinkage can develop in the main vascular trunk and why non-AMD CNV causes are more benign and can be stabilized longer with fewer injections.
In the literature, only Ong et al. demonstrated inactive CNV associated with ODD in 1 case with OCTA in a series of 8 paediatric patients with CNV. A 12-year-old female patient first underwent injection and photodynamic therapy in the left eye; however, OCTA revealed two large calibre vessels and a lack of fine capillaries, anastomoses, and vessel loops, which was considered as inactive sub-RPE CNV lesion [23]. This is the first paper to report a case with an interesting change in CNV visualization after injection and followed up for a long term for both CNV and vascular density. On the other hand, studies on OCTA for ODD mainly focus on microvascular changes. Biçer and Atilla showed that macular vascular density located in different regions decreased both at  -hemi  123  130  121  129  122  142  Inferior-hemi  137  122  131  121  136  126  Nasal superior  111  107  116  102  117  110  Nasal inferior  107  97  100  91  108  98  Inferior nasal  145  146  142  141  149  147  Inferior temporal  195  173  202  182  211  184  Temporal inferior  108  74  87  78  82  80  Temporal superior  100  97  88  100  91  107  Superior temporal  176  179  173  185  174  201  Superior nasal  118  157  119  150  117  174 Cells are divided into two to show the results of the right and left eyes separately.     [27]. In our case, however, peripapillary capillary density evaluation revealed that the vascular density in the nasal quadrants was lower than those in the other quadrants, which was an interesting finding. However, changes that might be correlated in both RNFL and disc and macular vascular density measurements were not detected contrary to the abovementioned reports. In conclusion, it is possible to evaluate vascular density, presence of associated CNV, and response to injection with OCTA in patients with optic nerve diseases such as ODD. Nonetheless, there is a need for further studies to confirm the changes in CNV visualization like in our case with OCTA that has been increasingly used.

Disclosure
The authors have no proprietary or commercial interest in any materials discussed in this article.

Conflicts of Interest
The authors declare that they have no conflict of interest.