Aflibercept Anti-Vascular Endothelial Growth Factor Therapy in Vitrectomized Eyes with Neovascular Age-Related Macular Degeneration

Editor

Anti-vascular endothelial growth factor (VEGF) therapy has been shown to stabilize and improve vision in eyes treated for neovascular age-related macular degeneration (NVAMD) (Schmidt-Erfurth et al. 2014), but none of the patients in the clinical trials were noted to have a history of prior pars plana vitrectomy (PPV); and the clinical question often arises whether intravitreal anti-VEGF therapy would be less effective in vitrectomized eyes. Previous studies of intravitreal medications have shown that clearance rates are faster after previous vitrectomy (Gisladottir et al. 2009). Therefore, intravitreal anti-VEGF therapy could potentially be less effective in vitrectomized eyes due to decreased viscosity and increased fluid currents leading to a decreased half-life and access to receptors (Christoforidis et al. 2013). We report an observational case series of four patients who were previously vitrectomized unilaterally, subsequently developed treatment naïve, NVAMD and were treated with aflibercept to adequately control their disease.

We retrospectively reviewed 4 patients (3 female) who developed NVAMD after prior PPV, followed for 23.8±0.5 months (mean±standard deviation) and were treated with intravitreal aflibercept (0.5 mg/0.5 ml, Eylea, Regeneron, Tarrytown, NY) pro re nata (PRN) after an initial loading regimen of 3 monthly injections based on spectral domain optical coherence tomography (SD-OCT) and clinical findings of recurrence of disease, defined as the presence of cystoid macular edema or subretinal fluid on SD-OCT or hemorrhage on fundoscopy seen between January 2012 and January 2015 by a single physician (S.C.). Their clinical histories, best correct visual acuity (BCVA, in logMAR) and SD-OCT findings including central foveal thickness (CFT) and initial lesion type are summarized in Table 1.

Table 1.

Clinical Case Summary Information

Case #	Age at Initial NVAMD Diagnosis, Race, Sex, Eye	Type of Macular Surgery	Time Between PPV and NVAMD (months)	BCVA after PPV	BCVA at NVAMD Diagnosis	BCVA at 2 years	# of injections year 1	# of injections year 2	Follow-up Time on Aflibercept (months)	CNV Type	CFT Prior to Aflibercept (µm)	CFT at 2 years (µm)
1	81 WF, OD	None		0.0	0.0	0.0	5.0	2.0	24	Type 1	268	246
	81 WF, OS	Macular Pucker	117	0.0	0.60	0.18	9.0	9.0	24	Type 3 with sub-RPE	558	369
2	83 WF, OD	Macular Pucker	8	0.18	0.54	0.30	7.0	7.0	23	Type 1	363	336
3	93 WF, OS	Macular Pucker	13	0.18	0.40	0.10	8.0	10.0	24	Type 3 with vitelliform	457	395
4	83 WM, OD	Stage 3 Macular Hole	131	0.30	2.0	2.0	9.0	5.0	24	Type 3 with sub-RPE	420	195

# = number; BCVA = best corrected visual acuity in logMAR; CFT = central foveal thickness; CNV = choroidal neovascularization; F = female; M = male; NVAMD = neovascular age-related macular degeneration; OD = right eye; OS = left eye; PPV = pars plana vitrectomy; RPE = retinal pigment epithelium; W = Caucasian

Mean age was 85±5.4 years old (range 81–93). On average, patients required 7.9±1.0 aflibercept injections/year (range 6.8–8.7) to avoid recurrence. While on aflibercept treatment, BCVA improved from 0.88±0.74 (range 0.40–2.0) to 0.64±0.91 (range 0.10–2.0) at 2 years with an average improvement of 0.24±0.18 (range 0.0–0.43). CFT improved from 449.5±82.0 µm (range 363–558) to 323.8±89.2 µm (range 195–295) at 2 years with an average decrease of 125.8±96.0 µm (range 27–225). There was clinical and anatomical improvement except in Case 4, where atrophy developed in the macula.

Case 1 of an 81 year-old female who had undergone left eye (OS) PPV for a macular pucker 10 years prior, illustrates a patient that received bilateral aflibercept injections for treatment naïve NVAMD. At diagnosis OS, SD-OCT demonstrated a mature type 3 (retinal angiomatous proliferation) and she received aflibercept at a rate of 8.7 injections/year OS. Five months after diagnosis OS, her right eye (OD) developed a type 1 choroidal neovascularization (sub-retinal pigment epithelium, CNV) and required an average of 4.7 injections/year OD. Interestingly, the average frequency of PRN injections required for disease-control in Case 1’s non-vitrectomized OD eye was equivalent to that reported in the second year of the VIEW1/2 trials, which was approximately 4 injections/year (Schmidt-Erfurth et al. 2014); whereas the vitrectomized OS eye, required more frequent treatments.

After vitrectomy, the diffusion gradients and fluid currents of intravitreal medications have been shown to be significantly faster (Gisladottir et al. 2009; Christoforidis et al. 2013). Animal studies have shown conflicting results regarding the effectiveness of anti-VEGF therapy after vitrectomy, and the variation in results may be attributable to heterogeneity in methods used for measuring clearance (Christoforidis et al. 2013; Ahn et al 2014). Moreover, the validity of translating these results into the human context remains uncertain, given the different volumes of vitreous and potential differences in pharmacokinetics (Ahn et al 2014).

Evidence of anti-VEGF treatment for NVAMD in vitrectomized human eyes remains lacking within the literature. Recently, Hahn reported on a single case after prior macular translocation vitrectomy surgery for NVAMD and showed that monthly aflibercept injections were successful in treating recurrent CNV after an initial poor response to a single bevacizumab injection (Hahn 2014). In comparison to this report utilizing monthly injections, our series of 4 vitrectomized eyes showed that an average of 7.9 aflibercept injections per year was effective in controlling treatment naïve NVAMD over 2 years. Aflibercept has been shown to have about a 140 times higher VEGF-binding affinity compared to ranibizumab (Steward & Rosenfeld 2008). This increased binding affinity may be a major reason why VEGF-trap has a prolonged duration of action, and although the half-life of the drug may also be decreased in vitrectomized eyes, this difference may explain its effectivity in these patients.

Our study is limited by its uncontrolled nature; small sample size; the underlying CNV type may affect treatment response; and our study assumes that NVAMD arising in vitrectomized eyes is pathologically equivalent to non-vitrectomized eyes. Regardless, to our knowledge, this is the first reported series of treatment naïve, NVAMD patients that were initially treated with aflibercept in vitrectomized eyes. Future randomized controlled studies comparing NVAMD in non-vitrectomized and vitrectomized eyes with the different anti-VEGF medications may further elucidate the differences in therapeutic response.

Clinicians should be aware that vitrectomized eyes may require increased frequency of anti-VEGF therapy due to possible changes in drug availability and utilizing an agent with a higher binding affinity and consistent treatment intervals may provide clinical stability.