Conversion to Aflibercept For Chronic Refractory OrRecurrent Neovascular Age-Related Macular

Degeneration

YOSHIHIRO YONEKAWA, CHRISTOPHER ANDREOLI, JOHN B. MILLER, JOHN I. LOEWENSTEIN,LUCIA SOBRIN, DEAN ELIOTT, DEMETRIOS G. VAVVAS, JOAN W. MILLER, AND IVANA K. KIM

� PURPOSE: To explore the visual and anatomicoutcomes of patients with refractory or recurrent neovas-cular age-related macular degeneration (AMD) who wereconverted from bevacizumab and/or ranibizumab to afli-bercept.� DESIGN: Two-center, retrospective chart review.� METHODS: Treatment history, visual acuity (VA), andcentral macular thickness (CMT) on spectral-domainoptical coherence tomography were collected. Patientswere divided into ‘‘refractory’’ (persistent exudationdespite monthly injections) or ‘‘recurrent’’ (exudationsuppressed, but requiring frequent injections).� RESULTS: One hundred and two eyes of 94 patientswere included; 68 were refractory and 34 were recurrent.Eyes received a mean of 20.4 prior bevacizumab/ranibizu-mab injections and a mean of 3.8 aflibercept injections.Mean follow-up was 18 weeks. Mean VA was 20/50-1before conversion, 20/50-2 after 1 aflibercept injection(P [ .723), and 20/50D2 after the final injection(P [ .253). Subgroup analysis of refractory and recur-rent cases also showed stable VA. Of the refractory cases,mean CMT had improved after 1 injection (P < .001)and the final injection (P < .001). Intraretinal(P < .001) and subretinal (P < .001) fluid decreasedafter 1 injection, and the mean injection interval wasextended from 5.2 to 6.2 weeks (P[ .003). Of the recur-rent cases, mean CMT improved after 1 injection(P < .001) and the final injection (P < .001). Intrare-tinal (P [ .003) and subretinal (P [ .046) fluiddecreased after 1 injection, and the mean injectioninterval was extended from 7.2 to 9.5 weeks (P[ .001).� CONCLUSIONS: Converting patients with chronicneovascular AMD to aflibercept results in stabilizedvision and improved anatomic outcomes, while allowinginjection intervals to be extended. (Am J Ophthalmol

Accepted for publication Mar 21, 2013.From Retina Service, Department of Ophthalmology, Massachusetts

Eye and Ear Infirmary, Harvard Medical School (Y.Y., C.A., J.B.M.,J.I.L., L.S., D.E., D.G.V., J.W.M., I.K.K); and Retina Service,Department of Ophthalmology, Harvard Vanguard Medical Associates(C.A.), Boston, Massachusetts.

Inquiries to Ivana K. Kim, Associate Professor of Ophthalmology,Harvard Medical School, Retina Service, Massachusetts Eye and EarInfirmary, 243 Charles Street, Boston, MA 02114; e-mail: Ivana_Kim@meei.harvard.edu

0002-9394/$36.00http://dx.doi.org/10.1016/j.ajo.2013.03.030

� 2013 BY ELSEVIER INC.

2013;-:-–-. � 2013 by Elsevier Inc. All rightsreserved.)

THE INTRODUCTION OF INTRAVITREAL ANTI–

vascular endothelial growth factor (VEGF) agentsgreatly improved the visual prognosis for patients

with neovascular age-related macular degeneration(AMD).1 The efficacy of monthly ranibizumab (Lucentis;Genentech, South San Francisco, California, USA), anaffinity-matured anti-VEGF antibody fragment that bindsall VEGF-A isoforms, was demonstrated in the MARINA2

(Minimally Classic/Occult Trial of the Anti-VEGF Anti-body Ranibizumab In the Treatment of NeovascularAMD) and ANCHOR3 (Anti-VEGF Antibody for theTreatment of Predominantly Classic Choroidal Neovascu-larization in AMD) trials. Bevacizumab (Avastin; Genen-tech) is the parent molecule of ranibizumab and is USFood and Drug Administration (FDA) approved for thetreatment of several systemic malignancies. Its intravitrealuse has become widely adopted as an off-label, cost-effec-tive alternative to ranibizumab, with recent randomizedcontrolled data demonstrating comparable efficacy in treat-ing neovascular AMD.4,5

However, despite the encouraging results of multipletrials and clinical experience, many patients requirecontinued monthly injections because of recurrent exuda-tion, and others have suboptimal responses or developtachyphylaxis.6 The recent introduction of aflibercept(Eylea; Regeneron, Tarrytown, New York, USA) providesan alternative mechanism of VEGF blockade. Afliberceptis a soluble decoy receptor, engineered by fusing VEGFreceptor (VEGFR) 1 and VEGFR2 to the Fc portion ofhuman immunoglobulin G-1 (IgG-1), allowing it to bindall isoforms of VEGF-A, VEGF-B, and placental growthfactor (PlGF).7,8 Aflibercept has been shown to havehigher binding affinities compared to ranibizumab andbevacizumab, which allows a less frequent dosing scheduleof every 8 weeks after 3 initial monthly injections.9 Its effi-cacy in patients with newly diagnosed neovascular AMDwas demonstrated in the VIEW trials (VEGF Trap-Eye:Investigation of Efficacy and Safety in Wet AMD [VIEW1, VIEW 2]), which were 2 parallel phase III trials.10

Since the FDA approval of aflibercept for neovascularAMD, many practitioners are converting patients from

1ALL RIGHTS RESERVED.

TABLE 1. Conversion to Aflibercept for Chronic NeovascularAge-Related Macular Degeneration: Patient Characteristics

Demographics

Patients (eyes) 94 (102)

Mean age (range) 79.6 (57-93)

Women, n (%) 58 (61.1%)

White race, n (%) 90 (94.7%)

Right eye, n (%) 49 (48.0%)

bevacizumab or ranibizumab to aflibercept with the goal ofextending treatment intervals or achieving betterresponses for refractory cases. In the current study, weexamined the short-term visual and anatomic outcomesof patients with neovascular AMD refractory to, or recur-rent on, bevacizumab and/or ranibizumab, who wereswitched to aflibercept injections. We also examinedwhether aflibercept conversion could extend the injectionintervals to decrease injection burden.

Refractory, n 68

Recurrent, n 34

Previous injections,a mean (median; range) 20.4 (18; 3-65)

Aflibercept injections, mean (median; range) 3.8 (3.5; 1-8)

aIncludes bevacizumab and/or ranibizumab.

METHODS

THIS STUDY IS A RETROSPECTIVE, INTERVENTIONAL,

noncomparative case series of consecutive patients treatedat the Massachusetts Eye and Ear Infirmary and HarvardVanguard Medical Associates. Subjects were identified inour electronic medical records as those with neovascularAMD who were previously treated with ranibizumab and/orbevacizumab and then converted to aflibercept betweenDecember 1, 2011, and July 31, 2012. Exclusion criteriawere concomitant visually significant ocular pathology(such as vitreous hemorrhage), insufficient clinical records,fewer than 3 previous anti-VEGF injections, and lack offollow-up after conversion to aflibercept. The study protocolwas approvedby the InstitutionalReviewBoards of theMassa-chusetts Eye and Ear Infirmary (IRB protocol #12-073H) andHarvard Pilgrim Health Care (IRB protocol #376548). Theprotocols allowed retrospective data collection of patientswith neovascular AMD who were treated with intravitrealaflibercept injections. The study complied with the HealthInsurance Portability and Accountability Act of 1996 andconformed to the tenets of the Declaration of Helsinki.

Treatment schedules, re-treatment schedules, and injec-tion methods were at the discretion of individual retinaspecialists. Local and systemic adverse events were assessedat every clinic visit. Visual acuities were converted fromSnellen notation to the logarithm of the minimal angleof resolution (logMAR) scale. Spectral-domain opticalcoherence tomography (SDOCT) was performed by Cirrus(Carl Zeiss Meditec, Dublin, California, USA), Spectralis(Heidelberg Engineering, Heidelberg, Germany), orRTVue (Optovue, Fremont, California, USA) SDOCTsystems. Individual patients were followed using the sameSDOCT system. Central macular thickness (CMT) valueswere obtained using the integrated software. Baselinepreconversion OCTs were defined as those performed onthe day of aflibercept conversion. The injection intervalsfor prior anti-VEGF therapies and for aflibercept werecalculated from the most recent interval for the respectivetreatments for each patient. Patients were divided into2 groups: ‘‘refractory’’ (persistent intraretinal and/or subre-tinal fluid despite monthly injections) or ‘‘recurrent’’(patients who responded well, but required frequentrepeated injections to maintain a dry macula). Subretinal

2 AMERICAN JOURNAL OF

fluid and intraretinal fluid were categorized into presenceof fluid, no fluid, or trace fluid, based on review of entirevolume scans. Trace fluid was defined as a few fine intrare-tinal cysts or noncentral thin slivers of subretinal fluid.Paired variables before and after aflibercept conversion

were analyzed using the Wilcoxon signed rank test.Nonparametric distribution was confirmed using histogramplots. All statistical tests were 2-tailed and significance wasdefined as P < .05. Missing data were not imputed. Statis-tical analysis was performed using Stata version 9.0 (Stata-Corp, LP, College Station, Texas, USA).

RESULTS

� PATIENT CHARACTERISTICS: One hundred patients(108 eyes) with neovascular AMD who were convertedfrom bevacizumab or ranibizumab to aflibercept were iden-tified. Six patients (6 eyes) were excluded because ofvitreous hemorrhage (1), corneal pathology (1), lack offollow-up (1), and use of ranibizumab only once before afli-bercept conversion in newly diagnosed patients (3). Theremaining cohort was composed of 94 patients(102 eyes). Patient characteristics at the time of afliberceptconversion are summarized in Table 1.

� TREATMENT CHARACTERISTICS: Sixty-eight eyes hadrefractory neovascular AMD, and 34 were recurrent. Ofthe refractory cases, 22 were previously treated with ranibi-zumab alone, 24 with bevacizumab alone, and 22 with bothranibizumab and bevacizumab. Of the recurrent cases, 26were previously treated with ranibizumab alone, 2 withbevacizumab alone, and 6 with both ranibizumab and beva-cizumab. The lengths of previous anti-VEGF treatmentwere 141.7 weeks overall, 126.2 weeks for refractorypatients, and 172.8 weeks for recurrent patients. Themost recent injection intervals are displayed in Table 2,which is a reflection of the clinical activity at the time ofconversion.

--- 2013OPHTHALMOLOGY

TABLE 2. Treatment Response After Converting to Aflibercept in Patients With Chronic Neovascular Age-RelatedMacular Degeneration

All P Refractory P Recurrent P

Mean visual acuity (logMAR)a

Before conversion (SD) 0.42 (0.30) 0.44 (0.33) 0.38 (0.23)

After 1 injection (SD) 0.44 (0.36) .723 0.46 (0.41) .897 0.39 (0.25) .778

Final (SD) 0.38 (0.27) .253b 0.38 (0.28) .215 0.38 (0.25) .811

Mean CMT (mm)

Before conversion (SD) 305.07 (80.65) 311.57 (77.83) 288.83 (86.86)

After 1 injection (SD) 274.05 (68.98) <.001 283.33 (66.97) <.001 250.83 (69.82) <.001

Final (SD) 276.20 (69.82) <.001 283.01 (68.73) <.001 260.97 (70.00) .001

Injection intervals (weeks)

Previous (SD) 5.86 (2.55) 5.18 (1.59) 7.21 (3.44)

Aflibercept (SD) 7.27 (2.94) <.001 6.15 (2.18) .003 9.47 (3.03) .001

CMT ¼ central macular thickness; logMAR ¼ logarithm of the minimal angle of resolution; SD ¼ standard deviation.aLogMAR values of 0.30, 0.40, and 0.50 are equivalent to 20/40, 20/50, 20/63 in Snellen notation, respectively.bSecond P values are also compared to values before aflibercept conversion.

In addition, 6 eyes also had a history of being treatedwith photodynamic therapy (PDT), 1 with thermal laser,and 2 with pegaptanib (Macugen; Eyetech, New York,New York, USA). Thirty-eight of 102 eyes (37.3%) werepreviously treated with enhanced doses of bevacizumab(2.00 mg) or ranibizumab (0.75 mg or 1.00 mg). Allpatients were treated using a pro re nata (PRN) and/ortreat-and-extend protocol. Conversion to aflibercept tookplace on average 7.0 weeks after the last bevacizumab orranibizumab injection for both groups combined, 5.7 weeksfor the refractory group, and 9.5 weeks for the recurrentgroup. The mean time between the first aflibercept injec-tion and the subsequent OCT was 5.2 weeks overall,5.0 weeks in refractory patients, and 5.7 weeks in recurrentpatients. Mean follow-up was 18.4 weeks after the firstinjection of aflibercept (17.6 weeks for refractory,19.9 weeks for recurrent).

� VISUAL OUTCOMES: Visual outcomes are summarized inTable 2. Mean baseline visual acuities before afliberceptconversion were approximately 20/50 in the combined,refractory, and recurrent groups. Mean visual acuityremained stable after 1 injection of aflibercept in all groups.There was a trend towards improved mean visual acuity inthe combined and refractory groups on the final visit, butthis was not statistically significant.

� ANATOMIC OUTCOMES: CMT was significantlydecreased after 1 injection and at the final visit in all groups(Figure 1; Table 2). The proportion of eyes with intrareti-nal and subretinal fluid decreased after 1 injection(Figure 2). When the intraretinal and subretinal fluid wasgraded from none to trace to present, there were statisti-cally significant improvements in the combined and refrac-tory groups (Table 3). In recurrent cases, the proportion ofpatients with intraretinal fluid decreased after 1 injection

VOL. -, NO. - CONVERSION TO AFLIBERCEPT FOR CH

and at the final visit, whereas the proportion with subreti-nal fluid decreased after 1 injection and trended lower atthe final visit, but lost statistical significance.Qualitatively, upon evaluation of all available OCT

studies after conversion, the anatomic response improvedin 61 of 67 refractory patients (91.0%) (1 patient did nothave a follow-up OCT) and was stable in 6 (9.0%), and nopatients worsened. All recurrent patients had continuedgood anatomic response.A tear of the retinal pigment epithe-lium (RPE) occurred after aflibercept conversion in 1 patient,but visual acuity remained stable and exudation improved.Trace subretinal hemorrhage developed in another patient,but both visual acuity and OCT improved. There were nocases of endophthalmitis. No systemic complications ordeaths were noted during the follow-up period.

� INJECTIONINTERVALS: The intravitreal injection inter-vals were able to be extended in all groups after convertingto aflibercept (Table 2). Overall the mean injectioninterval increased from 5.9 weeks on prior therapy to7.3 weeks with aflibercept (5.2-6.1 weeks for refractorygroup; 7.2-9.6 weeks for recurrent group). Of the refractorycases, the last aflibercept-aflibercept interval was 6 weeks orless in 45 eyes (66.2%), 7-9 weeks in 16 eyes (23.5%), and10 or more weeks in 7 eyes (10.3%). Of the recurrent eyes,the injection interval was <_6 weeks in 4 eyes (11.8%),7-9 weeks in 14 eyes (41.2%), and >_10 weeks in 16 eyes(47.1%). No eyes required shorter intervals than they didbefore conversion to aflibercept.

DISCUSSION

AFLIBERCEPT IS THEMOSTRECENTADDITIONTOTHEARMA-

mentarium of intravitreal anti-VEGF agents. The high

3RONIC MACULAR DEGENERATION

FIGURE 2. Anatomic response to aflibercept conversion forchronic neovascular age-related macular degeneration: propor-tion of eyes with intraretinal and subretinal fluid at baselineand after 1 injection of aflibercept.

FIGURE 1. Spectral-domain optical coherence tomography images of refractory neovascular age-related macular degeneration thatresponded to conversion to aflibercept. (Top row) A 78-year-old woman with neovascular age-related macular degeneration wastreated with 3 monthly injections of bevacizumab 1.25mg, and then 3monthly injection of bevacizumab 2.00mg, and then 7monthlyinjections of ranibizumab, with persistent subretinal fluid (Top left). The subretinal fluid completely resolved after 1 injection ofaflibercept (Top right), and her visual acuity improved from 20/80 to 20/40. (Bottom row) A 73-year-old man with neovascularage-related macular degeneration was treated with 12 monthly injections of ranibizumab 0.5 mg, and then 17 monthly injectionsof bevacizumab 1.25 mg, and then 2 monthly injections of ranibizumab 1.0 mg, for a total of 31 injections, with persistent subretinalfluid (Bottom left). The subretinal fluid completely resolved after 1 injection of aflibercept (Bottom right), and his visual acuityremained stable at 20/200.

binding affinity of the decoy receptor allows for bimonthlydosing after 3 initial monthly doses, as demonstrated by theVIEW trials.10 The properties of aflibercept that result inlonger duration of action could also theoretically lead toimproved efficacy in patients with suboptimal responsesto ranibizumab or bevacizumab.

In the current study, we examined the visual andanatomic response of such patients with refractory or recur-rent exudation who were converted to aflibercept. Theindications for initiating treatment with aflibercept were:(1) to elicit a better anatomic and possibly visual response;and/or (2) to extend the injection interval. Our datashowed that overall, visual acuity was maintained,

4 AMERICAN JOURNAL OF

anatomic responses improved, and injection intervalscould be slightly extended.The improved anatomic response was most evident in

the refractory group, where patients had persistent intrare-tinal and/or subretinal fluid despite monthly ranibizumaband/or bevacizumab. Both intraretinal and subretinal fluidimproved significantly after only 1 injection, and overallcontinued to show anatomic benefit until the last docu-mented visit. However, the anatomic gains did not trans-late into improved visual acuity during our follow-upperiod. This was not surprising, because the patients inthis series had received a mean of 20 prior injections(average of 20.4 in the refractory group and 22.8 in therecurrent group) with continuous or recurrent exudation.Out of both groups, 24 had 30 or more injections, 10 had40 or more, 5 had 50 or more, and 1 refractory patienthad 65 prior injections. Longer-term follow-up afterconversion is necessary to provide a more accurate assess-ment of visual acuity outcomes.The main goal in converting to aflibercept in the recur-

rent cases was to extend the injection intervals. Thesepatients had the benefit of ‘‘complete’’ responses to previousinjections of ranibizumab and/or bevacizumab, but hadrecurrent exudation that required frequent, many timesmonthly, injections. The baseline CMT is elevated in therecurrent group, because the baseline OCT includedpatients on a PRN schedule that developed recurrentexudation at the time of aflibercept conversion. Afliberceptconversion allowed the injection intervals to be extendednot only in this group but also in the refractory cases.The mechanism for increased durability has been

--- 2013OPHTHALMOLOGY

TABLE 3. Changes in Intraretinal and Subretinal Fluid After Converting to Aflibercept in Eyes With Chronic Neovascular Age-RelatedMacular Degeneration

All (Eyes) Refractory (Eyes) Recurrent (Eyes)

No Tr Yes P No Tr Yes P No Tr Yes P

Intraretinal fluid

Before conversion 30 1 71 16 1 51 14 0 20

After 1 injectionb 53 4 24 <.001 33 4 20 <.001 20 0 4 .003

Finalb 54 4 40 <.001a 34 3 30 <.001 20 1 10 .005

Subretinal fluid

Before conversion 53 1 48 30 1 37 23 0 11

After 1 injectionb 57 7 17 <.001 35 7 15 <.001 22 0 2 .046

Finalb 65 8 25 <.001 41 8 18 <.001 24 0 7 .180

Tr ¼ trace.aSecond P values are also compared to baseline.bTotal number of eyes do not add up to 102 because 4 patients did not have optical coherence tomographies (OCTs) after conversion to

aflibercept, and 21 did not have OCTs after 1 injection but did so subsequently.

attributed to aflibercept’s high VEGF binding affinity. Anin vitro study from Regeneron showed that aflibercepthad 94 and 119 times the binding affinity to VEGF-A165

compared to ranibizumab and bevacizumab, respectively.9

Aflibercept was also 33-71 times more potent than ranibi-zumab at inhibiting VEGF-A-induced receptor activationin human embryonic kidney (HEK) 293 cell lines trans-fected to express VEGFR1 or VEGFR2. However, it isalso notable that despite the increased potency of afliber-cept, the mean injection interval was only able to beextended by a mean of less than 2 weeks in this series ofpatients.

With respect to comparative efficacy of different anti-VEGF agents, a few groups have retrospectively reportedon the conversion to ranibizumab from other therapies.In a series of 102 eyes with neovascular AMD that werepreviously treated with bevacizumab or pegaptanib, Ehlersand associates found that 29 eyes (28.4%) gained morethan 3 lines after ranibizumab conversion.11 Their superiorvisual outcomes were likely attributable to shorter diseaseduration: on average, subjects received 2.9 bevacizumabor 1.8 pegaptanib injections prior to conversion, in contrastto our average of 20.4. In Stepien and associates’ review of84 eyes converted from bevacizumab to ranibizumab, theydid not find a benefit in visual acuity or injectionfrequency.12 Their cohort had, on average, 4.7 bevacizu-mab injections. OCT data were not included in either ofthe studies, but in Bhatnagar and associates’ review of189 patients previously treated with multiple modalitiesbefore conversion to ranibizumab, central macular thick-ness improved.13 Patients previously treated with bevacizu-mab did not experience an improvement in visual acuity,whereas those treated with PDT and pegaptanib did. Theabove studies collectively suggest that conversion frombevacizumab to ranibizumab may yield improvements inanatomy, but not visual acuity, in chronic AMD. Eyes

VOL. -, NO. - CONVERSION TO AFLIBERCEPT FOR CH

with potential for gaining vision appear to be those previ-ously treated with older, less efficacious agents such asPDT and pegaptanib, or nonresponders who wereconverted early.Tachyphylaxis refers to diminished responses to medica-

tion after repeated administration, and is a recognizedphenomenon complicating treatment of neovascularAMD with anti-VEGF agents.6,14–16 The enhancedanatomic response seen with aflibercept in our previouslytreated patients may represent stronger clinical activityof aflibercept or tachyphylaxis to the prior chronictreatments. Previous reports have described switchinganti-VEGF agents to counteract tachyphylaxis.6 Othermethods of treating challenging cases have also beenproposed, including higher doses17 and increasedfrequency18 of bevacizumab or ranibizumab. TheHARBORStudy demonstrated no difference in visual acuity resultsbetween ranibizumab 2.0 mg and 0.5 mg for treatment-naı̈ve neovascular AMD at 1 year (Suner IJ, et al. IOVS2012;53:ARVO E-Abstract 3677). However, Brown andassociates recently reported a prospective study (Super-dose Anti-VEGF [SAVE] trial) examining ranibizumab2.0 mg in the treatment of patients with neovascularAMD unresponsive to monthly 0.5 mg ranibizumab. Theydemonstrated a þ3.3 letter improvement (P ¼ .001)and�33.1mmof CMT (P¼ .01) after 3 monthly injectionsof 2.0 mg ranibizumab in 86 participants.19 Baseline visualacuities were comparable to our cohort, at approximately20/40, but the SAVE cohort received a mean of 10.4 priorinjections, compared to 20.4 in our refractory patients.Our retrospective analysis did not include protocol refrac-tions, but we did observe a small VA improvement in ourrefractory patients (þ0.06 logMAR), which is of similarmagnitude to the gains seen in the SAVE trial. Neverthe-less, the differences in study methodologies do not allowfor a conclusive comparison.

5RONIC MACULAR DEGENERATION

Our data suggest that converting patients with refractoryor recurrent neovascular AMD to aflibercept results inmaintenance of vision and improved anatomic outcomes,while allowing for slightly increased injection intervals.Strengths of the study design include the fact that eachsubject served as his or her own historical control andwas treated by the same retina specialist with the samere-treatment criteria (before and after conversion).

However, there are also several limitations to the currentstudy in addition to its retrospective nature. This studylacked standardized visual acuity measurements and refrac-tions, and re-treatment criteria varied among treatingphysicians. Our cohort represents a chronically treatedpopulation with an average of 20 prior injections over thecourse of 142 weeks, likely ensuring maximal treatmentresponse to therapy prior to conversion. However, the retro-spective nature did not allow standardization of the clinicalmanagement prior to aflibercept conversion. Additionally,

6 AMERICAN JOURNAL OF

the first injection of aflibercept was given on average 7weeksafter the last injection of ranibizumab or bevacizumab,which is beyond the optimal bioavailabilities of theseagents, whereas the first follow-up OCT after afliberceptconversion was at 5.2 weeks on average, which is still withinaflibercept’s peak bioavailability. This may have theoreti-cally biased the difference in CMT at the first follow-upvisit. Also, macular thickness measurements were auto-mated and not corrected for segmentation errors. The base-line CMTs are largely reflective of prior CMTs because ofthe chronicity of our patients, but the lack of more precon-version OCT data is another limitation of this study.Finally, the follow-up interval is limited, so the resultsreflect short-term outcomes of aflibercept conversion. It isyet unknown whether visual acuity will continue to remainstable, and whether the positive anatomic outcomes seen inthe study will persist. Further studies are warranted to deter-mine whether the benefits are sustained.

ALL AUTHORSHAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSUREOF POTENTIAL CONFLICTS OF INTEREST.The authors report the following disclosures: D. Eliott: consultant for Acucela, ArcticDx, Genentech, Glaukos, Ophthotech, Regeneron, Salutaris; I.K.Kim: consultant for ArticDx, Sequenom, SalutarisMD, Genzyme, Regeneron, and Genentech, and research support from Genentech; J.W. Miller: boardmember for Alcon (ended May 2011), consultant for Alcon and KalVista. The Massachusetts Eye and Ear Infirmary has an ownership interest in 3 U.S.patents directed to the use of verteporfin (Quadra Logic Technologies). In addition, the Massachusetts Eye and Ear Infirmary has ownership interest incertain patent applications directed to the selective destruction of subretinal choroidal neovasculature for the treatment of macular degeneration and otherdisorders. The Massachusetts Eye and Ear Infirmary receives royalties as a result of these patents and patent applications and J.W.M. receives a share of thesame in accordance with the Massachusetts Eye and Ear Infirmary’s institutional Patent Policy and Procedures, which includes royalty-sharing provisions.The authors indicate no funding support. Contributions of authors: design and conduct of the study (Y.Y., C.A., I.K.K.); collection, management, analysis,and interpretation of data (Y.Y., C.A., J.B.M., J.I.L., L.S., D.E., D.G.V., J.W.M., I.K.K.); and preparation, review, or approval of themanuscript (Y.Y., C.A.,J.B.M., J.I.L., L.S., D.E., D.G.V., J.W.M., I.K.K.).

REFERENCES

1. Andreoli CM, Miller JW. Anti-vascular endothelial growthfactor therapy for ocular neovascular disease. Curr OpinOphthalmol 2007;18(6):502–508.

2. Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab forneovascular age-related macular degeneration. N Engl J Med2006;355(14):1419–1431.

3. Brown DM, Michels M, Kaiser PK, Heier JS, Sy JP,Ianchulev T. Ranibizumab versus verteporfin photodynamictherapy for neovascular age-related macular degeneration:two-year results of the ANCHOR study. Ophthalmology

2009;116(1):57–65.4. Chakravarthy U, Harding SP, Rogers CA, et al. Ranibizumab

versus bevacizumab to treat neovascular age-related maculardegeneration: one-year findings from the IVAN randomizedtrial. Ophthalmology 2012;119(7):1399–1411.

5. Martin DF, Maguire MG, Fine SL, et al. Ranibizumab andbevacizumab for treatment of neovascular age-relatedmacular degeneration: two-year results. Ophthalmology 2012;119(7):1388–1398.

6. Gasperini JL, Fawzi AA, Khondkaryan A, et al. Bevacizumaband ranibizumab tachyphylaxis in the treatment of choroidalneovascularisation. Br J Ophthalmol 2012;96(1):14–20.

7. Browning DJ, Kaiser PK, Rosenfeld PJ, Stewart MW. Afliber-cept for age-related macular degeneration: a game-changer orquiet addition? Am J Ophthalmol 2012;154(2):222–226.

8. Stewart MW. Aflibercept (VEGF Trap-eye): the newest anti-VEGF drug. Br J Ophthalmol 2012;96(9):1157–1158.

9. Papadopoulos N, Martin J, Ruan Q, et al. Binding andneutralization of vascular endothelial growth factor (VEGF)and related ligands by VEGF Trap, ranibizumab and bevacizu-mab. Angiogenesis 2012;15(2):171–185.

10. Heir JS, Brown DM, Chong V, et al. Intravitreal aflibercept(VEGF trap-eye) in wet age-related macular degeneration.Ophthalmology 2012;119(12):2537–2548.

11. Ehlers JP, Spirn MJ, Shah CP, et al. Ranibizumab for exuda-tive age-related macular degeneration in eyes previouslytreated with alternative vascular endothelial growth factorinhibitors. Ophthalmic Surg Lasers Imaging 2010;41(2):182–189.

12. Stepien KE, Rosenfeld PJ, Puliafito CA, et al. Comparison ofintravitreal bevacizumab followed by ranibizumab for thetreatment of neovascular age-related macular degeneration.Retina 2009;29(8):1067–1073.

13. Bhatnagar P, Spaide RF, Takahashi BS, et al. Ranibizumab fortreatment of choroidal neovascularization secondary to age-related macular degeneration. Retina 2007;27(7):846–850.

14. Forooghian F, Cukras C, Meyerle CB, Chew EY, Wong WT.Tachyphylaxis after intravitreal bevacizumab for exudativeage-related macular degeneration. Retina 2009;29(6):723–731.

15. Keane PA, Liakopoulos S, Ongchin SC, et al. Quantitativesubanalysis of optical coherence tomography after treatment

--- 2013OPHTHALMOLOGY

with ranibizumab for neovascular age-related macular degen-eration. Invest Ophthalmol Vis Sci 2008;49(7):3115–3120.

16. Schaal S, Kaplan HJ, Tezel TH. Is there tachyphylaxis tointravitreal anti-vascular endothelial growth factor pharma-cotherapy in age-related macular degeneration? Ophthal-mology 2008;115(12):2199–2205.

17. Fung AT, Kumar N, Vance SK, et al. Pilot study to evaLuatethe role of high-dose rAnibizumab 2.0 mg in the managementof neovascular age-related macular degeneration in patientswith perSistent/recurrenT macular fluid <30 days following

VOL. -, NO. - CONVERSION TO AFLIBERCEPT FOR CH

treatment with intravitreal anti-VEGF therapy (the LASTStudy). Eye (Lond) 2012;26(9):1181–1187.

18. Stewart MW, Rosenfeld PJ, Penha FM, et al. Pharmacokineticrationale for dosing every 2 weeks versus 4 weeks with intravi-treal ranibizumab, bevacizumab, andaflibercept (vascular endo-thelial growth factor Trap-eye). Retina 2012;32(3):434–457.

19. Brown DM, Chen E,Mariani A,Major JC Jr. Super-dose anti-VEGF (SAVE) trial: 2.0 mg intravitreal ranibizumab forrecalcitrant neovascular macular degeneration-primary endpoint. Ophthalmology 2013;120(2):349–354.

7RONIC MACULAR DEGENERATION

Biosketch

Yoshihiro Yonekawa, MD, graduated from Weill Cornell Medical College and is currently a resident at the Massachusetts

Eye and Ear Infirmary.

7.e1 --- 2013AMERICAN JOURNAL OF OPHTHALMOLOGY

Biosketch

Ivana K. Kim, MD, is Associate Professor of Ophthalmology at Harvard Medical School and a member of the Retina

Service at Massachusetts Eye and Ear Infirmary. She graduated from Harvard Medical School and completed her

ophthalmology residency and vitreoretinal fellowship at the Massachusetts Eye and Ear Infirmary. Her clinical practice

includes surgical and medical retina, with a focus on age-related macular degeneration and uveal melanoma. She is

actively involved in clinical trials for retinal disease as well as studies of age-related macular degeneration genetics and

molecular genetics of uveal melanoma.

VOL. -, NO. - 7.e2CONVERSION TO AFLIBERCEPT FOR CHRONIC MACULAR DEGENERATION