Intravitreal Aflibercept (VEGF Trap-Eye) in
`Wet Age-related Macular Degeneration
`
`
`
`Jeffrey S. Heier, MD,' David M. Brown, MD,” Victor Chong, MD,? Jean-Francois Korobelnik, MD,?
`Peter K. Kaiser, MD,” Quan Dong Nguyen, MD,° Bernd Kirchhof, MD,’ Allen Ho, MD,®
`Yuichiro Ogura, MD,” GeorgeD. Yancopoulos, MD,PhD, '0 Neil Stahl, MD,10 Robert Vitti, MD,'°
`Alyson J. Berliner, MD, PhD,'° Yuhwen Soo, PhD,!° Majid Anderesi, MD,'' Georg viperach, MD,!!
`Bernd Sommerauer, PhD,'! Ruyupert Sandbrink, MD, PhD,"!"!? Christian Simader, MD
`Ursula Schmidt-Erfurth, MD,””
`for the VIEW 1 and VIEW 2 Study Groups*
`
`Twosimilarly designed, phase-3 studies (VEGF Trap-Eye: Investigation of Efficacy and Safety in
`Objective:
`Wet AMD [VIEW 1, VIEW 2]) of neovascular age-related macular degeneration (AMD) compared monthly and
`every-2-month dosing of intravitreal aflibercept injection (VEGF Trap-Eye; Regeneron, Tarrytown, NY, and Bayer
`HealthCare, Berlin, Germany) with monthly ranibizumab.
`Design: Double-masked, multicenter, parallel-group, active-controlled, randomizedtrials.
`Participants: Patients (n = 2419) with active, subfoveal, choroidal neovascularization (CNV) lesions (or
`juxtafoveal lesions with leakage affecting the fovea) secondary to AMD.
`Intervention: Patients were randomized to intravitreal aflibercept 0.5 mg monthly (0.5q4), 2 mg monthly
`(2q4), 2 mg every 2 monthsafter 3 initial monthly doses (2q8), or ranibizumab 0.5 mg monthly (Rq4).
`Main Outcome Measures: The primary end point was noninferiority (margin of 10%) of the aflibercept
`regimens to ranibizumabin the proportion of patients maintaining vision at week 52 (losing <15 letters on Early
`Treatment Diabetic Retinopathy Study [ETDRS] chart). Other key end points included changein best-corrected
`visual acuity (BCVA) and anatomic measures.
`Results: All aflibercept groups were noninferior andclinically equivalent to monthly ranibizumabfor the
`primary end point (the 2q4, 0.5q4, and 2q8 regimens were 95.1%, 95.9%, and 95.1%, respectively, for VIEW
`1, and 95.6%, 96.3%, and 95.6%, respectively, for VIEW 2, whereas monthly ranibizumab was 94.4% in both
`studies). In a prespecified integrated analysis of the 2 studies, all aflibercept regimens werewithin 0.5 letters
`of the reference ranibizumab for mean change in BCVA; all aflibercept regimens also produced similar
`improvements in anatomic measures. Ocular and systemic adverse events were similar across treatment
`groups.
`Intravitreal aflibercept dosed monthly or every 2 monthsafter3 initial monthly doses produced
`Conclusions:
`similar efficacy and safety outcomes as monthly ranibizumab. These studies demonstrate that aflibercept is an
`effective treatment for AMD, with the every-2-month regimen offering the potential to reduce the risk from
`monthly intravitreal injections and the burden of monthly monitoring.
`Financial Disclosure(s): Proprietary or commercial disclosure may be found after the references.
`Ophthalmology 2012;119:2537-2548 © 2012 by the American Academy of Ophthalmology.
`
`=) “Group memberslisted online in Appendix 17 (http://aaojournal.org).
`
`Age-related macular degeneration (AMD) is a leading
`cause ofvision loss and blindness in industrialized coun-
`tries.' The most severe vision loss occurs in the neovas-
`cular (or wet) form of AMD, involving choroidal neo-
`vascularization (CNV) and associated retinal edema.
`Early treatments for CNV (laser ablation, photodynamic
`therapy with verteporfin), although clearly better than no
`treatment at all, decreased severe vision loss rather than
`truly stabilizing vision or resulting in clinically signifi-
`cant
`improvements in visual acuity.2~+ The suggestion
`that vascular endothelial growth factor (VEGF) might be
`driving the CNV and associated edema seen in AMDled
`to a paradigm shift with the success of the first anti-
`VEGF therapy, pegaptanib sodium.°° Monthly intravit-
`
`injections of 0.5 mg ranibizumab, a humanized
`real
`monoclonal antibody fragment
`that blocks VEGF, not
`only prevent vision loss in most patients but also lead to
`significant visual gain in approximately one-third.”* The
`risk of rare but serious adverse events resulting from the
`intravitreal procedure, together with the significant bur-
`den of making monthly visits to their retinal specialist,
`have led to extensive efforts to decrease injection and
`monitoring frequency. However,fixed quarterly®!° or “as
`needed” (pro re nata [PRN]) dosing regimens,!!*!? with-
`out requiring monthly monitoring visits, were not effec-
`tive at maintaining vision.
`The Comparison of AMD Treatments Trials (CATT)!*
`recently compared monthly ranibizumab with monthly
`
`© 2012 by the American Academy of Ophthalmology
`Published by Elsevier Inc.
`
`ISSN 0161-6420/12/$—see front matter
`http://dx.doi.org/10.1016/j.ophtha.2012.09.006
`
`2537
`
`CELLTRION - EXHIBIT 1018
`
`CELLTRION - EXHIBIT 1018
`
`

`

`Ophthalmology Volume 119, Number 12, December 2012
`
`tion of patients who had fluid-free retinas on optical
`bevacizumab,as well as with PRN regimensthat required
`coherence tomography (OCT). Although CIs were not
`monthly monitoring visits during which treatment deci-
`provided for monthly and PRN regimens, switching from
`sions primarily were made on the basis of anatomic
`monthly to PRN regimens in the second year of the
`criteria. Monthly bevacizumab resulted in mean best-
`CATTresulted in a significant worsening of BCVA and
`corrected visual acuity (BCVA) gains(8.0 letters) similar
`retinal thickness, as well as a significant decrease in the
`to those for monthly ranibizumab (8.5 letters), whereas
`proportion of patients without retinal fluid.'* The “alter-
`PRN ranibizumab yielded a mean BCVA gain of 1.7
`native treatments to Inhibit VEGFin Age-related choroi-
`letters less than that of the monthly standard (with a
`dal Neovascularization” (IVAN) study also found that the
`confidence interval [CI] extending to 4.7 letters below)
`mean foveal retinal thickness and the percentage of pa-
`that achieved noninferiority, and PRN bevacizumab
`tients with fluorescein leakage were significantly higher
`yielded a mean BCVAgain 2.6 letters below the monthly
`with the PRN regimen compared with the monthly regi-
`standard (with a CI extending to 5.9 letters below) that
`did not achieve noninferiority. In the CATT, monthly—men.!> In the HARBORstudy (Invest Ophthalmol Vis
`bevacizumab and both PRN regimens weresignificantly
`Sci 2012;53:E-Abstract 3677), PRN regimensof both the
`worse than monthly ranibizumab in termsof the propor-
`approved 0.5 mg dose and the higher 2 mg dose of
`
`Per Protocol Set: 285
` Analyzed:
`Per Protocol Set: 265
`fanwzumab
`1 (0.3%)
`
`VIEW 1
`
`Assessed
`for
`eligibility
`(n=2063)
`
`Randomized
`(n=1217)
`
`Excluded (n=846)
`
`A
`
`Discontinuation all IAl groups:
`Total:
`Adverse event:
`Death:
`Lost to follow-up:
`Protocol violation:
`Treatmentfailure:
`Withdrawalof consent:
`Other:
`
`65 (7.1%)
`12 (1.3%)
`10 (1.1%)
`10 (1.1%)
`4 (0.4%)
`4 (0.4%)
`20 (2.2%)
`5 (0.5%)
`
`Al
`2q4 (n=304)
`
`IAl
`0.594 (n=304)
`
`Al
`2q8 (n=303)
`
`Analyzed:
`Safety Analysis Set: 304
`Full Analysis Set: 304
`
`Safety Analysis Set: 304
`Full Analysis Set: 301
`Per Protocol Set: 270
`
`Analyzed:
`Safety Analysis Set: 303
`Full Analysis Set: 301
`
`Analyzed:
`Safety Analysis Set: 304
`Full Analysis Set: 304
`Per Protocol Set: 269
`
`Discontinuation Ranibizumab:
`Total:
`Adverse event:
`Death:
`Lost to follow-up:
`Protocol violation:
`Treatmentfailure:
`Withdrawalof consent:
`Other:
`
`22 (7.2%)
`4 (1.3%)
`3 (1.0%)
`1 (0.3%)
`3 (1.0%)
`0
`10 (3.3%)
`
`Figure 1. Flowcharts describing treatmentallocation and patient disposition in VIEW 1 (A) and VIEW 2 (B). In both VIEW 1 and VIEW 2 studies, the
`most commonreason for patients to be screened but not randomized wasineligibility based on angiographic characteristics as identified by the reading
`center. The second most common reason was visual acuity out of range. Discontinuations are those that occurred from the study. Two milligrams
`intravitreal aflibercept every 2 months (2q8) dosing was performed after 3 initial monthly doses. The numbers of patients who prematurely discontinued
`study medicationin the 2q4, 0.5q4, 2q8, and Rq4 groups were 16 (5.3%), 30 (9.9%), 30 (9.9%), and 27 (8.8%), respectively, in VIEW 1; and 37 (11.8%),
`45 (14.5%), 33 (10.5%), and 33 (10.9%), respectively, in VIEW 2. In VIEW 1, 1089 patients were included in the per protocol set (PPS), with 92.6%
`to 96.1% completing week-52 visual acuity assessment. A total of 128 patients were not included in the PPS for the following reasons (in order of
`occurrence): missed 2 consecutive injections before ninth injection, major protocol deviation, received <9 injections, had <9 assessments, no baseline
`assessments, no post-baseline assessments. In VIEW 2, 1081 patients were included in the PPS with 95.9% to 97.8% completing week-52 visual acuity
`assessment. A total of 159 patients were not included in the PPS for the following main reasons: missed 2 consecutive injections before ninth injection,
`major protocol deviation, received <9 injections, had <9 assessments, no baseline assessments, no post-baseline assessments, unmasking by investigator
`or Global Pharmacovigilance. 0.5q4 = 0.5 mg IAI monthly; 2q4 = 2 mg IAI monthly; 2q8 = 2 mg IAI every 2 months after 3 initial monthly doses;
`IAI = intravitreal aflibercept injection.
`
`2538
`
`

`

`Heier et al
`
`+ Intravitreal Aflibercept for Wet AMD
`
`action in the eye,!° allowing for less frequent dosing, as
`supported by early clinical trials.!*:?° In this article, we
`report the first-year results of 2 phase 3 studies compar-
`ing intravitreal aflibercept, monthly or every 2 months,
`with monthly ranibizumab.
`
`ranibizumab did not achieve noninferiority compared
`with monthly ranibizumab, with the 0.5 mg PRN regimen
`yielding a mean BCVA gain 2.0 letters below the
`monthly standard (with a CI extending to 4.5 letters
`below). Of note, just like the CATT PRN regimens, the
`HARBOR PRN regimens still depended on monthly
`monitoring visits. Thus,
`there remains a need for new
`therapies that will provide equivalent efficacy and ana-
`tomic disease control
`to monthly ranibizumab, while
`reducing the risk of monthly injections and the burden of
`mandatory monthly monitoring visits.
`(previously
`Intravitreal aflibercept
`injection (IAI)
`The “VEGFTrap-Eye: Investigation of Efficacy and Safety in Wet
`knownin the scientific literature as WEGF Trap-Eye,
`AMD”studies (VIEW 1 and VIEW 2) were similarly designed,
`Regeneron, Tarrytown, NY, and Bayer HealthCare, Ber-
`prospective, double-masked, multinational, parallel-group, active-
`lin, Germany) is a soluble decoy receptor fusion pro-
`controlled, randomized clinical trials. The investigators from the
`tein'®!7 that
`is specifically purified and formulated for
`VIEW 1 and VIEW2studies are listed in Appendix 1, available
`intraocular injection. Intravitreal aflibercept at doses of
`at http://aaojournal.org. Patients in VIEW 1 (registered at www.
`clinicaltrials.gov on July 31, 2007; NCT00509795. Accessed Au-
`0.5 mg and 2 mg provided the most robust outcomesin
`gust 8, 2012) were randomizedat 154sites in the United States and
`the Clinical Evaluation of Antiangiogenesis in the Retina
`Canada. Patients in VIEW 2 (registered at www.clinicaltrials.gov
`Intravitreal Trial Phase 2 (CLEAR-IT 2) study after 4
`on March 12, 2008; NCT00637377. Accessed August 8, 2012)
`monthly administrations followed by PRN dosing to
`were randomized at 172 sites in Europe, the Middle East, Asia-
`week 52.'® The binding affinity of intravitreal aflibercept
`Pacific, and Latin America; the last patient in both studies com-
`to VEGF is substantially greater than that of bevaci-
`pleted 52 weeks in September 2010. The study protocols were
`zumab or ranibizumab.!” The greater affinity could trans-
`approved byinstitutional review boards or ethics committees for
`late into a higher efficacy or, as predicted by a math-
`each clinical site; all participants provided written informed con-
`ematic model,
`into a substantially longer duration of
`sent. All the US study sites complied with the Health Insurance
`
`Materials and Methods
`
`Study Design
`
`
`
`Discontinuation all IAl groups:
`Total:
`Adverse event:
`Death:
`Lost to follow-up:
`Protocolviolation:
`Treatmentfailure:
`Withdrawalof consent:
`Other:
`
`98 (10.7%)
`23 (2.5%)
`6 (0.7%)
`5 (0.5%)
`2 (0.2%)
`2 (0.2%)
`39 (4.3%)
`21 (2.3%)
`
`VIEW 2
`
`
`Analyzed:
`
`
`
`Safety Analysis Set: 309
`
` 2q4 (n=313)
`Full Analysis Set: 309
`
`
`Per Protocol Set: 274
`
`
`Analyzed:
`IAI
`Safety Analysis Set: 297
`
`
`Per Protocol Set: 268
`Full Analysis Set: 296
`0.5q4 (n=311)
`
`
`Assessed
`
`
`for
`Randomized
`
`eligibility
`(n=1240)
`Analyzed:
`(n=2031)
`Safety Analysis Set: 307
`
`
`2q8 (n=313)
`Full Analysis Set: 306
`Per Protocol Set: 270
`
`
`
`
`
`Excluded (n=791)
`
`
`
`Figure 1. (Continued.)
`
`Analyzed:
`Ranibizumab
`
`Safety Analysis Set: 291
`(n=303)
`Full Analysis Set: 291
`
`Per Protocol Set: 269
`
`Discontinuation Ranibizumab:
`Total:
`Adverse event:
`Death:
`Lost to follow-up:
`Protocolviolation:
`Treatmentfailure:
`Withdrawalof consent:
`Other:
`
`27 (8.9%)
`2 (0.7%)
`1 (0.3%)
`4 (1.3%)
`2 (0.7%)
`0
`11 (3.6%)
`7 (2.3%)
`
`2539
`
`

`

`Ophthalmology Volume 119, Number 12, December 2012
`
`Portability and Accountability Act. The 52-week outcomes are
`reported.
`
`Participants
`Inclusion and exclusion criteria were designed to maintain
`constancy with the pivotal trials for the reference drug ranibi-
`zumab, consistent with regulatory guidelines for noninferiority
`studies, and included (1) age ⱖ50 years with active subfoveal
`CNV lesions (any subtype) secondary to AMD; juxtafoveal
`lesions with leakage affecting the fovea also were allowed; (2)
`CNV comprising at least 50% of total lesion size; and (3)
`BCVA between 73 and 25 Early Treatment Diabetic Retinop-
`athy Study chart (ETDRS) letters (20/40 –20/320 Snellen equiv-
`alent). Patients with prior treatment for AMD (including an
`investigational agent or anti-VEGF therapy) in the study eye
`were excluded. Eligibility was determined using fluorescein
`angiography at the reading center. Complete eligibility criteria
`are shown in Appendix 2 (available at http://aaojournal.org).
`
`Treatment Groups and Randomization
`Patients were randomized in a 1:1:1:1 ratio to the following
`regimens: 0.5 mg aflibercept every 4 weeks (0.5q4); 2 mg
`aflibercept every 4 weeks (2q4); 2 mg aflibercept every 8 weeks
`(2q8) after 3 injections at week 0, 4, and 8 (to maintain
`masking, sham injections were given at the interim 4-week
`visits after week 8); or 0.5 mg ranibizumab every 4 weeks
`(Rq4). Consecutively enrolled patients were assigned to treat-
`ment groups on the basis of a predetermined central random-
`ization scheme with balanced allocation, managed by an inter-
`active voice response system.
`
`End Points and Statistical Analyses
`The primary end point analysis, noninferiority margins, and
`definition of “clinical equivalence” were established in discus-
`sion with the Food and Drug Administration (FDA) (as part of
`a Special Protocol Assessment), European Medicines Agency,
`Pharmaceutical and Medical Device Agency and other regula-
`tory authorities, with the intent of maintaining constancy with
`the previous ranibizumab pivotal trials7,8 and preserving the
`majority of the treatment effect demonstrated in these trials.
`The primary end point analysis was noninferiority of the intra-
`vitreal aflibercept regimens to ranibizumab in the proportion of
`patients maintaining vision at week 52 (losing ⬍15 ETDRS
`letters; per protocol data set) in each study. A noninferiority
`margin of 10% in the individual studies was chosen to preserve
`approximately two-thirds of the ranibizumab effect for preven-
`tion of moderate vision loss (loss of ⬍15 letters) demonstrated
`in pivotal ranibizumab studies,7,8 using the 2 CI approach. The
`FDA suggested that a margin of 5% could determine clinical
`equivalence. Thus, the margin of 10% was used for assessing
`noninferiority, and the margin of 5% was used for assessing
`clinical equivalence. The prespecified analysis plan also in-
`cluded a prospectively planned integrated analysis combining
`the 2 VIEW studies; in this integrated analysis, the European
`Medicines Agency/Committee for Medicinal Products for Hu-
`man Use requested a noninferiority margin of 7%. In the
`individual studies, the primary end point was assessed by a
`prespecified hierarchical testing sequence of noninferiority to
`ranibizumab with the sequence of aflibercept 2q4, 0.5q4, and
`then 2q8 to control the 5% (4.9% for VIEW 1) overall type I
`error while maintaining a 5% significance level (4.9% for
`
`2540
`
`VIEW 1) for each individual comparison (see Appendices 3 and
`4 for details of the statistical analysis, available at http://
`aaojournal.org). If all aflibercept groups demonstrated noninfe-
`riority to ranibizumab for the primary end point, additional
`comparisons with ranibizumab were prespecified regarding the
`secondary end points, also using a hierarchical testing sequence
`in which each secondary end point was tested for superiority of
`aflibercept over ranibizumab. Prespecified secondary efficacy
`variables compared baseline and 52-week data regarding mean
`change in BCVA; gaining ⱖ15 letters; change in total National
`Eye Institute 25-Item Visual Function Questionnaire (NEI
`VFQ-25) score; and change in CNV area on fluorescein angiog-
`raphy. Anatomic measures included retinal thickness and per-
`sistent fluid as assessed by OCT. Change in BCVA also was
`assessed as part of the prospectively planned prespecified inte-
`grated analysis combining the 2 studies.
`The full analysis set included all randomized patients who
`received any study medication and had a baseline and at least 1
`post-baseline BCVA assessment. The per protocol set (PPS)
`included all patients in the full analysis set who (1) received at
`least 9 doses of study drug and attended at least 9 scheduled
`visits during the first year, (2) had not missed 2 consecutive
`injections before administration of the ninth injection (per pa-
`tient), and (3) did not have major protocol violations. Sham
`injections were counted as doses administered for the purpose
`of defining the PPS. The PPS included patients who discontin-
`ued the study because of treatment failure, without a major
`protocol deviation, at any time during the first 52 weeks (even
`if they met points 1 and 2 above). These patients were consid-
`ered nonresponders for the primary end-point analysis. The last
`observation carried forward (LOCF) approach was used to
`impute missing values. When indicated, the robustness of anal-
`ysis results was assessed by using the observed case or com-
`pleters’ data. A completer was defined as a patient who received
`treatment for at least 9 months and had efficacy data for at least
`9 months during the 52 weeks of study. The missing values for
`completers also were imputed using the LOCF approach.
`
`Schedule of Visits and Assessments
`
`Patients were examined on the day of treatment initiation and
`every 4 weeks thereafter through 52 weeks, as well as 1 week after
`first treatment for safety assessment (subsequent safety assess-
`ments occurred by telephone). Each 4-week visit included BCVA
`assessment and anterior/posterior segment examination (with in-
`traocular pressure determination) before injection (active or sham)
`and posterior segment examination with intraocular pressure de-
`termination 30 to 60 minutes after injection. For the 2q8 treatment
`group, no treatment decisions were made at the interim monthly
`visits. The NEI VFQ-25 assessment occurred at screening and
`weeks 12, 24, 36, and 52. Adverse events were recorded at every
`visit.
`
`Imaging Assessments
`
`Fundus photography and fluorescein angiography were performed
`at screening and weeks 24 and 52, and evaluated by an indepen-
`dent center (Digital Angiography Reading Center, New York).
`Optical coherence tomography was performed using time domain
`Stratus machines (Carl Zeiss Meditec, Jena, Germany) and eval-
`uated by an independent center (VIEW 1: OCT Reading Center at
`Duke, Durham, NC; VIEW 2: Vienna Reading Center, Austria).
`Visual acuity examiners were certified to ensure consistent mea-
`surement of BCVA. In VIEW 1, OCT was performed at screening,
`at the treatment initiation visit, and at weeks 4, 12, 24, 36, and 52
`
`
`
`
`

`

`Heier et al
`
`䡠 Intravitreal Aflibercept for Wet AMD
`
`(and was optional at the investigators’ discretion at other study
`visits). In VIEW 2, OCT was performed at every study visit. Areas
`of visible CNV (classic or occult) were identified when angio-
`graphic analyses showed evidence of late leakage or pooling
`of dye.
`
`Masking
`
`Patients were masked as to treatments. An unmasked investigator
`performed the study drug or sham injection. An unmasked inves-
`tigator also was responsible for the receipt, tracking, preparation,
`destruction, and administration of study drug, as well as safety
`assessments both pre- and post-dose. A separate masked physician
`assessed adverse events and supervised the masked assessment of
`efficacy. All other study site personnel were masked to treatment
`assignment by separating study records or masked packaging.
`Optical coherence tomography technicians and visual acuity ex-
`aminers remained masked relative to treatment assignment. Intra-
`vitreal aflibercept and sham kits were packaged identically. Lu-
`centis (Genentech Inc, South San Francisco, CA) was obtained
`commercially but only prepared and delivered by unmasked per-
`sonnel at the sites.
`
`Results
`
`Patient Disposition, Baseline Characteristics, and
`Exposure
`The disposition of patients is shown in Figure 1A-B. In VIEW 1,
`1217 patients were randomized, with 91.1% to 96.4% of patients
`completing 52 weeks. In VIEW 2, 1240 patients were randomized,
`with 88.1% to 91.1% completing 52 weeks. Baseline demograph-
`ics and disease characteristics were evenly balanced among all
`treatment groups (Table 1). The mean number of active injections
`received by patients in all monthly treatment arms, which were
`scheduled to receive 13 monthly injections, was 12.1 to 12.5 in
`VIEW 1 and 12.2 to 12.4 in VIEW 2. The aflibercept every-2-
`month groups, scheduled to receive 3 initial monthly injections
`followed by 5 active injections over the next 10 months, received
`an average of 7.5 active injections in VIEW 1 and in VIEW 2.
`
`Primary End Point Analysis
`In both studies, the proportion of patients maintaining vision was
`similar among all treatment groups in the prespecified per-protocol
`analysis and the full analysis set (Table 2). All aflibercept groups
`achieved statistical noninferiority compared with monthly ranibi-
`zumab, with the CIs of the difference between ranibizumab and
`
`Table 1. Patient Demographics and Baseline Characteristics
`
`Ranibizumab
`0.5q4
`
`VIEW 1
`Intravitreal Aflibercept
`0 5q4
`
`2q4
`
`2q8
`
`304
`78.2⫾7.6
`
`304
`77.7⫾7.9
`
`301
`78.4⫾8.1
`
`301
`77.9⫾8.4
`
`296 (97.4)
`1 (0.3)
`0
`7 (2.3)
`
`295 (97.0)
`1 (0.3)
`3 (1.0)
`5 (1.6)
`
`291 (96.7)
`0
`5 (1.7)
`5 (1.7)
`
`287 (95.3)
`1 (0.3)
`4 (1.3)
`9 (3.0)
`
`Ranibizumab
`0 5q4
`
`291
`73.0⫾9.0
`
`213 (73.2)
`1 (0.3)
`60 (20.6)
`17 (5.8)
`
`VIEW 2
`Intravitreal Aflibercept
`0 5q4
`
`2q4
`
`2q8
`
`309
`74.1⫾8.5
`
`296
`74.7⫾8.6
`
`306
`73.8⫾8.6
`
`226 (73.1)
`0
`67 (21.7)
`16 (5.2)
`
`219 (74.0)
`1 (0.3)
`61 (20.6)
`15 (5.1)
`
`217 (70.9)
`2 (0.7)
`69 (22.5)
`18 (5.9)
`
`132 (43.4)
`172 (56.6)
`54.0⫾13.4
`
`110 (36.2)
`194 (63.8)
`55.2⫾13.2
`
`134 (44.5)
`167 (55.5)
`55.6⫾13.1
`
`123 (40.9)
`178 (59.1)
`55.7⫾12.8
`
`122 (41.9)
`169 (58.1)
`53.8⫾13.5
`
`133 (43.0)
`176 (57.0)
`52.8⫾13.9
`
`149 (50.3)
`147 (49.7)
`51.6⫾14.2
`
`131 (42.8)
`175 (57.2)
`51.6⫾13.9
`
`4.3% (13)
`
`4.9% (15)
`
`6.3% (19)
`
`6.6% (20)
`
`2.7% (8)
`
`2.6% (8)
`
`5.4% (16)
`
`3.3% (10)
`
`6.53⫾5.2
`
`6.59⫾5.1
`
`6.49⫾4.5
`
`6.57⫾5.1
`
`7.59⫾5.3
`
`8.25⫾5.8
`
`7.70⫾5.3
`
`7.75⫾5.5
`
`82 (27.0)
`
`87 (28.6)
`
`81 (26.9)
`
`71 (23.6)
`
`70 (24.1)
`
`72 (23.3)
`
`80 (27.0)
`
`88 (28.8)
`
`101 (33.2)
`115 (37.8)
`15 (4.9)
`
`105 (34.5)
`110 (36.2)
`13 (4.3)
`
`97 (32.2)
`121 (40.2)
`17 (5.6)
`
`110 (36.5)
`118 (39.2)
`17 (5.6)
`
`104 (35.7)
`116 (39.9)
`20 (6.9)
`
`112 (36.2)
`123 (39.8)
`15 (4.9)
`
`103 (34.8)
`113 (38.2)
`11 (3.7)
`
`106 (34.6)
`110 (35.9)
`14 (4.6)
`
`6.99⫾5.5
`
`6.98⫾5.4
`
`6.95⫾4.7
`
`6.89⫾5.2
`
`8.01⫾5.7
`
`8.72⫾6.1
`
`8.17⫾5.5
`
`8.22⫾5.9
`
`315.3⫾108.3
`
`313.6⫾103.4
`
`313.2⫾106.0
`
`324.4⫾111.2
`
`325.9⫾110.9
`
`334.6⫾119.8
`
`326.5⫾116.5
`
`342.6⫾124.0
`
`71.8⫾17.2
`
`70.4⫾16.6
`
`71.1⫾17.8
`
`69.6⫾16.8
`
`72.9⫾19.1
`
`70.3⫾19.4
`
`74.0⫾18.2
`
`71.3⫾19.1
`
`N (full analysis set)
`Age, yrs (mean ⫾ SD)
`Race
`White
`Black
`Asian
`Other
`Sex
`Men, n (%)
`Women, n (%)
`Baseline ETDRS BCVA
`(mean ⫾ SD)
`Proportion of patients with
`ⱖ20/40 BCVA, % (n)
`CNV area, mm2
`(mean ⫾ SD)
`Lesion type
`Predominantly classic,
`n (%)
`Minimally classic, n (%)
`Occult, n (%)
`Patients with juxtafoveal
`lesions, n (%)
`Lesion size, mm2
`(mean ⫾ SD)
`Central retinal thickness,
`␮m (mean ⫾ SD)
`Baseline NEI VFQ-25
`scores (mean ⫾ SD)
`
`0.5q4 ⫽ 0.5 mg monthly; 2q4 ⫽ 2 mg monthly; 2q8 ⫽ 2 mg every 2 months after 3 initial monthly doses; BCVA ⫽ best-corrected visual acuity; CNV ⫽
`choroidal neovascularization; ETDRS ⫽ Early Treatment Diabetic Retinopathy Study; NEI VFQ-25 ⫽ National Eye Institute 25-Item Visual Functioning
`Questionnaire; SD ⫽ standard deviation.
`
`2541
`
`
`
`
`

`

`Ophthalmology Volume 119, Number 12, December 2012
`
`Ranibizumab
`0.5q4
`
`269
`94.4% (254)
`
`304
`93.8% (285)
`
`304
`8.1⫾15.3
`
`Table 2. Prespecified Efficacy
`
`VIEW 1
`Intravitreal Aflibercept
`0.5q4
`
`270
`95.9% (259)
`
`301
`95.0% (286)
`
`2q4
`
`285
`95.1% (271)
`
`304
`95.1% (289)
`
`2q8
`
`265
`95.1% (252)
`
`301
`94.4% (284)
`
`304
`10.9⫾13.8
`3.15 (0.92 to 5.37)
`
`301
`6.9⫾13.4
`⫺0.80 (⫺3.03 to 1.43)
`
`301
`7.9⫾15.0
`0.26 (⫺1.97 to 2.49)
`
`30.9% (94)
`
`37.5% (114)
`
`24.9% (75)
`
`30.6% (92)
`
`6.58 (⫺0.98 to 14.14)
`
`⫺6.00 (⫺13.17 to 1.16)
`
`⫺0.36 (⫺7.74 to 7.03)
`
`⫺4.2⫾5.6
`
`⫺4.6⫾5.5
`
`⫺3.5⫾5.3
`
`⫺3.4⫾6.0
`
`⫺0.33 (⫺1.04 to 0.38)
`
`0.71 (⫺0.01 to 1.42)
`
`0.86 (0.15–1.58)
`
`4.9⫾14.0
`
`6.7⫾13.5
`
`4.5⫾11.9
`
`5.1⫾14.7
`
`1.28 (⫺0.73 to 3.28)
`
`⫺0.67 (⫺2.69 to 1.35)
`
`⫺0.60 (⫺2.61 to 1.42)
`
`⫺116.8⫾109.0
`
`⫺116.5⫾98.4
`
`⫺115.6⫾104.1
`
`⫺128.5⫾108.5
`
`63.6% (171)
`
`64.8% (184)
`
`56.7% (148)
`
`63.4% (168)
`
`Primary end point
`N (PPS)
`Proportion maintaining vision (losing
`⬍15 ETDRS letters), % (n)
`N (full analysis set)
`Proportion maintaining vision (losing
`⬍15 ETDRS letters, LOCF), % (n)
`Secondary end points
`N (full analysis set)
`Change in ETDRS BCVA (mean ⫾ SD)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Proportion gaining ⱖ15 ETDRS letters,
`% (n)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Change in CNV area, mm2
`(mean ⫾ SD)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Change in total NEI VFQ-25 score
`(mean ⫾ SD)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Exploratory end point
`Change in central retinal thickness, ␮m
`(mean ⫾ SD)
`Post hoc end point†
`Proportion with dry retina (absence of
`cystic intraretinal edema and
`subretinal fluid on OCT), % (n)
`
`0.5q4 ⫽ 0.5 mg monthly; 2q4 ⫽ 2 mg monthly; 2q8 ⫽ 2 mg every 2 months after 3 initial monthly doses; BCVA ⫽ best-corrected visual acuity;
`aflibercept injection; LOCF ⫽ last observation carried forward; LS ⫽ least-squares; NEI VFQ-25 ⫽ National Eye Institute 25-Item Visual
`*95.1% CI for VIEW 1.
`†Observed case.
`
`each aflibercept group within the prespecified 10% margin (Fig 2),
`and the point estimates of the differences in means favoring the
`aflibercept groups in all cases. All the aflibercept regimens also
`met the prespecified 7% noninferiority margin in the prespecified
`integrated analysis combining the 2 VIEW studies, as well as the
`prespecified 5% margin for clinical equivalence compared with
`ranibizumab in the individual VIEW studies. Moreover, the results
`of multiple imputation analyses were consistent with those using
`the LOCF.
`
`Mean Changes in Best-Corrected Visual Acuity
`and Other Visual Acuity End Points
`The mean change in BCVA was a clinically important secondary
`end point in both studies. On the basis of the hierarchical testing
`sequence, only the aflibercept 2q4 group was statistically superior
`to ranibizumab, and only in VIEW 1, with a gain of ⫹10.9 versus
`⫹8.1 letters (Table 2). Small numeric differences between treat-
`ment groups in one study at any given timepoint were not repro-
`duced in the other study, suggesting that they reflected random
`variability even in groups of this size (Fig 3A, B); this interpreta-
`tion was supported by a prespecified integrated analysis that com-
`bined the 2 studies (Fig 3C), showing similar visual acuity scores
`
`across the entire 52-week study for all treatment groups. All
`groups behaved similarly in this integrated analysis (Fig 3C), with
`rapid increases in mean visual acuity after the first
`injection
`followed by incremental gains that were durable and maintained
`through week 52. Regardless of whether the analysis was by
`LOCF, by multiple imputations, by assessing completers, or by
`using actual observed data, intravitreal aflibercept dosed every 2
`months achieved a mean visual acuity score within 0.3 letters of
`monthly ranibizumab in the integrated analysis, with a CI of less
`than 2 letters (Fig 3C, inset).
`In both studies, the secondary end point of proportions of
`patients gaining ⱖ15 ETDRS letters from baseline to week 52 was
`similar in all treatment groups (Table 2), as were other exploratory
`categoric measures of visual outcome (Appendix 5, available at
`http://aaojournal.org). Likewise, vision-related quality of life, as-
`sessed by the change of total score of the NEI VFQ-25, improved
`in all groups in both studies (Table 2).
`
`Key Anatomic Measures
`In both studies, all groups demonstrated a comparable decrease
`in the secondary end point of change in area of active CNV
`
`2542
`
`
`
`
`

`

`Heier et al
`
`䡠 Intravitreal Aflibercept for Wet AMD
`
`Outcomes at Week 52
`
`Primary end point
`N (PPS)
`Proportion maintaining vision (losing
`⬍15 ETDRS letters), % (n)
`N (full analysis set)
`Proportion maintaining vision (losing
`⬍15 ETDRS letters, LOCF), % (n)
`Secondary end points
`N (full analysis set)
`Change in ETDRS BCVA (mean ⫾ SD)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Proportion gaining ⱖ15 ETDRS letters,
`% (n)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Change in CNV area, mm2
`(mean ⫾ SD)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Change in total NEI VFQ-25 score
`(mean ⫾ SD)
`LS mean difference between IAI and
`ranibizumab (95% CI)*
`Exploratory end point
`Change in central retinal thickness, ␮m
`(mean ⫾ SD)
`Post hoc end point†
`Proportion with dry retina (absence of
`cystic intraretinal edema and
`subretinal fluid on OCT), % (n)
`
`Ranibizumab
`0.5q4
`
`269
`94.4% (254)
`
`291
`94.8% (276)
`
`291
`9.4⫾13.5
`
`VIEW 2
`Intravitreal Aflibercept
`0.5q4
`
`268
`96.3% (258)
`
`296
`95.3% (282)
`
`2q4
`
`274
`95.6% (262)
`
`309
`94.5% (292)
`
`2q8
`
`270
`95.6% (258)
`
`306
`95.4% (292)
`
`309
`7.6⫾12.6
`⫺1.95 (⫺4.10 to 0.20)
`
`296
`9.7⫾14.1
`⫺0.06 (⫺2.24 to 2.12)
`
`306
`8.9⫾14.4
`⫺0.90 (⫺3.06 to 1.26)
`
`34.0% (99)
`
`29.4% (91)
`
`34.8% (103)
`
`31.4% (96)
`
`⫺4.57 (⫺12.02 to 2.88)
`
`0.78 (⫺6.91 to 8.46)
`
`⫺2.65 (⫺10.18 to 4.88)
`
`⫺4.2⫾5.9
`
`⫺6.0⫾6.1
`
`⫺4.2⫾6.1
`
`⫺5.2⫾5.9
`
`⫺1.18 (⫺1.98 to ⫺0.38)
`
`0.17 (⫺0.63 to 0.97)
`
`⫺0.73 (⫺1.53 to 0.07)
`
`6.3⫾14.8
`
`4.5⫾15.0
`
`5.1⫾13.7
`
`4.9⫾14.7
`
`⫺2.79 (⫺4.90 to ⫺0.68)
`
`⫺0.93 (⫺3.07 to 1.20)
`
`⫺1.95 (⫺4.07 to 0.17)
`
`⫺138.5⫾122.2
`
`⫺156.8⫾122.8
`
`⫺129.8⫾114.8
`
`⫺149.2⫾119.7
`
`60.4% (162)
`
`80.3% (220)
`
`63.9% (170)
`
`71.9% (197)
`
`CNV ⫽ choroidal neovascularization; CI ⫽ confidence interval; ETDRS ⫽ Early Treatment Diabetic Retinopathy Study; IAI ⫽ intravitreal
`Functioning Questionnaire; OCT ⫽ optical coherence tomography; PPS ⫽ per protocol set; SD ⫽ standard deviation.
`
`(Table 2). Likewise, all aflibercept groups in both studies had
`reductions in central retinal
`thickness similar to those for
`monthly ranibizumab as assessed by OCT, with a large and
`rapid reduction evident by week 4 (with retinal thickness ap-
`proaching normal levels) that was maintained to week 52 (Table
`2, Fig 4). Minor fluctuations in central retinal thickness were
`seen in the 2q8 group after sham injections in the VIEW 2
`study; these fluctuations attenuated over time, starting at 17 ␮m
`and decreasing to 8 ␮m over the year, with no apparent negative
`impact on visual acuity outcomes.
`Because of the inability of other regimens in the CATT13 to
`match the retinal thickness and retinal fluid improvements seen
`with monthly ranibizumab, a post hoc analysis was performed to
`determine the percentage of patients who had fluid-free retinas,
`which were defined, on OCT, by the absence of both cystic
`intraretinal edema and subretinal fluid. All intravitreal aflibercept
`groups were similar to the monthly ranibizumab group in terms of
`this end point, with numerically higher percentages of dry retinas
`seen in the 2q4 and 2q8 regimens largely driven by VIEW 2 (Table
`2; Appendix 6, available at http://aaojournal.org). Integrated anal-
`ysis combining both studies for proportions of patients with dry
`retinas for ranibizumab and the aflibercept regimens of 2q4, 0.5q4,
`and 2q8 showed percentages of 62.0%, 72.4%, 60.3%, and 67.7%,
`respectively.
`
`Figure 2. Difference in proportions of patients who maintained vision (losing
`⬍15 Early Treatment Diabetic Retinopathy Study [ETDRS] letters) at week 52 in
`the V