Randomized Trial Evaluating Ranibizumab
`Plus Prompt or Deferred Laser or
`Triamcinolone Plus Prompt Laser for
`Diabetic Macular Edema
`
`The Diabetic Retinopathy Clinical Research Network*
`Writing Committee: Michael J. Elman, MD; Lloyd Paul Aiello, MD, PhD; Roy W. Beck, MD, PhD; Neil M.
`Bressler, MD; Susan B. Bressler, MD; Allison R. Edwards, MS; Frederick L. Ferris III, MD; Scott M. Friedman,
`MD; Adam R. Glassman, MS; Kellee M. Miller, MPH; Ingrid U. Scott, MD, MPH; Cynthia R. Stockdale,
`MSPH; Jennifer K. Sun, MD, MPH. *The members of the DRCR Network who participated in this protocol are
`listed in Appendix 5.
`
`Objective: Evaluate intravitreal 0.5 mg ranibizumab or 4 mg triamcinolone combined with focal/grid laser
`compared with focal/grid laser alone for treatment of diabetic macular edema (DME).
`Design: Multicenter, randomized clinical trial.
`Participants: A total of 854 study eyes of 691 participants with visual acuity (approximate Snellen equiva-
`lent) of 20/32 to 20/320 and DME involving the fovea.
`Methods: Eyes were randomized to sham injection ⫹ prompt laser (n⫽293), 0.5 mg ranibizumab ⫹ prompt
`laser (n⫽187), 0.5 mg ranibizumab ⫹ deferred (ⱖ24 weeks) laser (n⫽188), or 4 mg triamcinolone ⫹ prompt laser
`(n⫽186). Retreatment followed an algorithm facilitated by a web-based, real-time data-entry system.
`Main Outcome Measures: Best-corrected visual acuity and safety at 1 year.
`Results: The 1-year mean change (⫾standard deviation) in the visual acuity letter score from baseline was
`significantly greater in the ranibizumab ⫹ prompt laser group (⫹9⫾11, P⬍0.001) and ranibizumab ⫹ deferred
`laser group (⫹9⫾12, P⬍0.001) but not in the triamcinolone ⫹ prompt laser group (⫹4⫾13, P⫽0.31) compared
`with the sham ⫹ prompt laser group (⫹3⫾13). Reduction in mean central subfield thickness in the triamcinolone
`⫹ prompt laser group was similar to both ranibizumab groups and greater than in the sham ⫹ prompt laser
`group. In the subset of pseudophakic eyes at baseline (n⫽273), visual acuity improvement in the triamcinolone
`⫹ prompt laser group appeared comparable to that in the ranibizumab groups. No systemic events attributable
`to study treatment were apparent. Three eyes (0.8%) had injection-related endophthalmitis in the ranibizumab
`groups, whereas elevated intraocular pressure and cataract surgery were more frequent in the triamcinolone ⫹
`prompt laser group. Two-year visual acuity outcomes were similar to 1-year outcomes.
`Conclusions:
`Intravitreal ranibizumab with prompt or deferred laser is more effective through at least 1 year
`compared with prompt laser alone for the treatment of DME involving the central macula. Ranibizumab as applied in
`this study, although uncommonly associated with endophthalmitis, should be considered for patients with DME and
`characteristics similar to those in this clinical trial. In pseudophakic eyes, intravitreal triamcinolone ⫹ prompt laser
`seems more effective than laser alone but frequently increases the risk of intraocular pressure elevation.
`Financial Disclosure(s): Proprietary or commercial disclosure may be found after
`the references.
`Ophthalmology 2010;117:1064 –1077 © 2010 by the American Academy of Ophthalmology.
`
`Macular edema is a frequent manifestation of diabetic retinop-
`athy and an important cause of impaired vision in individuals
`with diabetes.1–3 Focal/grid photocoagulation, the current stan-
`dard care for diabetic macular edema (DME), has been the
`mainstay of treatment since its benefit was demonstrated in the
`Early Treatment Diabetic Retinopathy Study (ETDRS) in
`1985.4 In a randomized, multicenter clinical trial, the Diabetic
`Retinopathy Clinical Research Network (DRCR.net) showed
`that focal/grid photocoagulation in eyes with center-involved
`DME and visual acuity ⱕ20/40 produces gradual visual acuity
`
`improvement of ⱖ2 lines in approximately one third of eyes
`after 2 years of follow-up, although approximately 20% of
`laser-treated eyes worsen by ⱖ2 lines.5 Thus, other treatment
`modalities, including anti-vascular endothelial growth factor
`(VEGF) therapy and steroids, alone or in combination with
`laser, are under investigation.
`The rationale for anti-VEGF therapy for DME is based on
`the observation that VEGF levels are increased in the retina
`and vitreous of eyes with diabetic retinopathy.6 Vascular en-
`dothelial growth factor has been demonstrated to increase
`
`1064
`
`© 2010 by the American Academy of Ophthalmology
`Published by Elsevier Inc.
`
`ISSN 0161-6420/10/$–see front matter
`doi:10.1016/j.ophtha.2010.02.031
`
`Celltrion Exhibit 1018
`Page 1
`
`

`

`DRCR Network Writing Committee 䡠 Laser-Ranibizumab-Triamcinolone RCT for DME
`
`vessel permeability in vivo possibly by increasing the phos-
`phorylation of tight junction proteins.7 Therefore, therapy that
`inhibits VEGF may represent a useful therapeutic modality that
`targets the underlying pathogenesis of DME. Pegaptanib
`(Macugen, Eyetech Pharmaceuticals, Palm Beach Gardens,
`FL) was the first anti-VEGF drug reported to have a favor-
`able effect on macular edema,8 although more recently, the
`anti-VEGF drugs ranibizumab (Lucentis, Genentech, South
`San Francisco, CA) and bevacizumab (Avastin, Genentech),
`among others, also have been evaluated for DME. Prior
`studies, which were small with short-term follow-up, have
`reported promising results.9 Intravitreal triamcinolone also
`was evaluated previously as treatment for DME in a ran-
`domized trial conducted by the DRCR.net.5 Although the
`data suggest that triamcinolone treatment was superior to
`the expected untreated course in the ETDRS, it was not
`superior to focal/grid photocoagulation.5
`The combination of intravitreal treatment (either triam-
`cinolone or an anti-VEGF drug) with focal/grid photocoag-
`ulation, theoretically, could be more effective than either
`treatment alone. The intravitreal treatment might rapidly
`reduce macular edema and lead to more rapid visual acuity
`improvement, whereas slower benefit accrues over time as a
`result of laser treatment. In addition, combined treatment
`could enhance the effect of focal/grid photocoagulation
`because the retina would be less edematous if laser treat-
`ment was administered some time after the intravitreal
`treatment reduced macular edema. Also, laser treatment
`theoretically could reduce the number of repeat intravitreal
`injections required to optimize the outcome of DME treat-
`ment. In a study of 86 eyes randomized to 4 mg intravitreal
`triamcinolone alone or followed by macular laser photoco-
`agulation, Kang et al10 reported that after 6 months visual
`acuity was better and more eyes had resolution of central
`edema with the combined treatment when compared with
`intravitreal triamcinolone without macular laser. Other stud-
`ies have shown greater mean visual acuity improvements at
`6 months using ranibizumab ⫹ laser, or ranibizumab alone,
`when compared with laser alone.9
`To determine whether anti-VEGF therapy alone or in
`combination with focal/grid laser, or intravitreal triamcino-
`lone combined with focal/grid laser, might result in im-
`proved outcomes compared with the standard treatment for
`DME of laser alone, the DRCR.net designed a clinical trial
`to evaluate 3 treatment modalities for DME in comparison
`with focal/grid photocoagulation: ranibizumab combined
`with prompt (within 1 week) focal/grid photocoagulation,
`intravitreal triamcinolone combined with prompt (within 1
`week) focal/grid photocoagulation, and intravitreal ranibi-
`zumab with focal/grid photocoagulation deferred for at least
`24 weeks. The study design also provided an opportunity to
`determine which regimen resulted in fewer treatments if
`safety and efficacy were comparable.
`
`Materials and Methods
`
`This phase 3 randomized, multicenter clinical trial was conducted
`by the DRCR.net at 52 clinical sites in the United States. The study
`adhered to the tenets of the Declaration of Helsinki. The protocol
`and informed consent forms were compliant with the Health In-
`
`surance Portability and Accountability Act and approved by mul-
`tiple institutional review boards. Each study participant gave writ-
`ten informed consent before participation in the study. Study
`oversight was provided by an independent data and safety moni-
`toring committee. The study was conducted under an Investiga-
`tional New Drug Application from the Food and Drug Adminis-
`tration. The study is listed on www.clinicaltrials.gov under
`identifier NCT00445003 (website registration date 03-06-2007),
`and the protocol is available on the DRCR.net website (www.drcr.
`net, date accessed January 1, 2010). Key aspects of the protocol
`pertinent to this article are summarized next.
`
`Study Population
`Eligible patients were at least 18 years old with type 1 or 2
`diabetes. The major eligibility criteria for a study eye included the
`following: (1) best-corrected Electronic-Early Treatment Diabetic
`Retinopathy Study (E-ETDRS Visual Acuity Test11) visual acuity
`letter score 78 to 24 (20/32–20/320), (2) definite retinal thickening
`due to DME on clinical examination involving the center of the
`macula assessed to be the main cause of visual loss, and (3) retinal
`thickness measured on time domain optical coherence tomography
`(OCT) ⱖ250 ␮m in the central subfield. Principal exclusion cri-
`teria included the following: (1) treatment for DME within the
`prior 4 months, (2) panretinal photocoagulation within the prior 4
`months or anticipated need for panretinal photocoagulation within
`the next 6 months, (3) major ocular surgery within the prior 4
`months, (4) history of open-angle glaucoma or steroid-induced
`intraocular pressure (IOP) elevation that required IOP-lowering
`treatment, and (5) IOP ⱖ25 mmHg. Patients were excluded if their
`systolic blood pressure was ⬎180 mmHg or diastolic blood pres-
`sure was ⬎110 mmHg, or if a myocardial infarction, other cardiac
`event requiring hospitalization, cerebrovascular accident, transient
`ischemic attack, or treatment for acute congestive heart failure
`occurred within 4 months before randomization. A patient could
`have 2 study eyes in the trial only if both were eligible at the time
`of study entry.
`
`Synopsis of Study Design
`After eligibility was determined and informed consent was ob-
`tained, study participants with 1 study eye were assigned randomly
`on the DRCR.net study website (using a permuted blocks design
`stratified by study eye visual acuity) with equal probability to 1 of
`4 treatment groups: (1) sham injection plus prompt (within 3–10
`days after injection) focal/grid photocoagulation (sham ⫹ prompt
`laser group), (2) 0.5 mg intravitreal ranibizumab plus prompt
`(within 3–10 days after injection) focal/grid photocoagulation
`(ranibizumab ⫹ prompt laser group), (3) 0.5 mg intravitreal ranibi-
`zumab with deferred (ⱖ24 weeks) focal/grid photocoagulation
`(ranibizumab ⫹ deferred laser group), and (4) 4 mg intravitreal
`triamcinolone plus prompt (within 3–10 days after injection) focal/
`grid photocoagulation (triamcinolone ⫹ prompt laser group). For
`study participants with 2 study eyes, the right eye was assigned
`randomly with equal probability to 1 of the 4 groups as indicated
`above. If the right eye was assigned to a treatment group other
`than the sham ⫹ prompt laser group, then the left eye was
`assigned to the sham ⫹ prompt laser group. If the right eye was
`assigned to the sham ⫹ prompt laser group, then the left eye
`was assigned randomly to 1 of the other 3 groups. Thus, there
`were more eyes in the sham ⫹ prompt laser group than in the
`other 3 groups.
`Follow-up was planned for 3 years, with the primary outcome
`at 1 year. During the first year, follow-up visits occurred every 4
`weeks (⫾1 week). Study participants in the 3 groups receiving
`laser were masked to treatment assignment through the primary
`
`1065
`
`Celltrion Exhibit 1018
`Page 2
`
`

`

`Ophthalmology Volume 117, Number 6, June 2010
`
`outcome visit, whereas the ranibizumab ⫹ deferred laser group
`was not masked. After the first year, visits occurred every 4 to 16
`weeks depending on the treatment group, disease course, and
`treatment administered. After a study participant completed the
`primary outcome visual acuity examination at 1 year, the study
`participant was made aware of his or her treatment group assign-
`ment and sham injections were discontinued. Visual acuity exam-
`iners and OCT technicians were masked to treatment group as-
`signment before and at the 1-year primary outcome visit.
`
`Examination Procedures
`At baseline and each follow-up visit, best-corrected visual acuity
`letter score was measured at 3 m by acertified examiner using an
`E-ETDRS Visual Acuity Test.11 The OCT images were obtained at
`baseline and each follow-up visit by a certified operator using the
`Zeiss Stratus OCT (OCT3) machine (Carl Zeiss Meditec, Inc.,
`Dublin, CA). Scans were 6 mm in length and included the 6-radial
`line fast macular scan pattern for quantitative measures and the
`cross-hair pattern (6 –12 o’clock and 9 –3 o’clock) for qualitative
`assessment of retinal morphology. All baseline OCT scans, annual
`follow-up scans with a standard deviation of the center point
`ⱖ10.0%, and scans from any visits in which the investigator
`suspected erroneous measurements because of the algorithm place-
`ment of the lines created by the OCT software that delineate the
`inner and outer aspects of the retina were sent to the Fundus
`Photograph Reading Center (University of Wisconsin, Madison)
`for grading. If the automated thickness measurements were judged
`by the Reading Center to be inaccurate on any submitted image,
`center point thickness was measured manually, and this value was
`used to impute a value for the central subfield based on a corre-
`lation of the 2 measures of 0.98 as published previously12 (20% of
`854 baseline scans were imputed and 1 scan was unable to be
`manually graded at baseline, and 2% of 10 849 follow-up scans
`were imputed and 22 [⬍1%] were unable to be manually graded
`during follow-up through 1 year). Manual grading of the baseline
`scans resulted in an imputed baseline central subfield value ⬍250
`␮m for 60 eyes (7%), which does not necessarily mean that the
`true thickness measurement is ⬍250 if measureable. Of note, 22
`(37%) of the 60 scans with an imputed central subfield thickness
`⬍250 ␮m were from 1 clinical site and represented 85% of the 26
`baseline scans from that site. All intent-to-treat results presented
`were similar when evaluated with exclusion of eyes from that
`clinical site (data not shown) and when evaluated with exclusion of
`eyes from any clinical site with a baseline central subfield thick-
`ness ⬍250 ␮m. Baseline OCT images also were assessed by the
`Reading Center for cystoid abnormalities and subretinal fluid.
`Additional testing at baseline and each follow-up visit included
`slit-lamp examination, measurement of IOP, and fundus examina-
`tion after pupil dilation. Standard ETDRS 7-field color stereo-
`scopic fundus photographs were obtained at baseline and 12
`months by a certified photographer and graded at the reading
`center for level of diabetic retinopathy.13 Hemoglobin A1c was
`measured at baseline. Any untoward medical occurrence, regard-
`less of whether the event was considered treatment related, was
`considered as an adverse event and recorded. Treatment of adverse
`events and proliferative diabetic retinopathy was at the discretion
`of the investigator.
`
`Treatment Protocol
`
`Overview. The treatment protocol (summarized in Appendix 1,
`available at http://aaojournal.org) included a baseline treatment
`followed by intravitreal study drug or sham injection retreatments
`every 4 weeks through the 12-week study visit. From the 16-week
`study visit and thereafter, a retreatment algorithm for study drug
`
`1066
`
`injections and sham injections (Appendices 2 and 3, available at
`http://aaojournal.org) was designed to require retreatments unless
`a study visit was deemed a ‘success’ (defined below and in Table
`1, available at http://aaojournal.org) at which point retreatment
`was at investigator discretion. From the 24-week study visit and
`thereafter retreatment was at investigator discretion if the study
`visit was deemed ‘no improvement’ (defined in Table 1, available
`at http://aaojournal.org). If retreatment with a study drug or sham
`injection was not given, ‘alternative treatment’ (defined in Table 1,
`available at http://aaojournal.org) was permitted only if a study eye
`met criteria for ‘failure’ or ‘futility’ (defined in Table 1, available
`at http://aaojournal.org). When retreatment with a study drug or
`sham injection was indicated, eyes assigned to one of the ranibi-
`zumab groups could receive ranibizumab as often as every 4
`weeks; eyes assigned to intravitreal triamcinolone could receive
`triamcinolone as often as every 16 weeks with sham injections as
`often as every 4 weeks in between triamcinolone injections; eyes
`assigned to sham ⫹ prompt laser could receive sham injections as
`often as every 4 weeks. A retreatment algorithm for focal/grid laser
`(Appendix 4, available at http://aaojournal.org) was designed to
`require retreatment if there was ‘edema involving the center of the
`macula’ or ‘edema threatening the center of the macula’ (defined
`in Table 1, available at http://aaojournal.org) and if ‘complete
`laser’ had not been given (defined in Table 1, available at http://
`aaojournal.org), provided that it had been at least 13 weeks since
`the last focal/grid laser application.
`Retreatment Algorithm System. Compliance with the details
`of the treatment protocol, which depended mainly on visual acuity
`and OCT measurements over time, was facilitated by a web-based,
`real-time data-entry system. At each follow-up visit, the system
`provided real-time feedback to the treating physician regarding
`whether treatment was required or at investigator discretion. If
`treatment was to be given, the system also provided feedback as to
`whether the treatment should be an intravitreal study drug or sham
`injection, whether focal/grid photocoagulation should be applied,
`and what the next follow-up interval should be.
`
`Statistical Methods
`Data are reported that were collected by the clinical sites from
`March 2007 to February 8, 2010. This includes at least 1-year
`follow-up for the entire study population and up to 2-year
`follow-up for participants enrolled early in the trial. Mean
`change in visual acuity from baseline to 1 year adjusted for
`baseline visual acuity was the primary outcome measure. The
`primary analysis consisted of 3 pairwise comparisons of the
`mean change in the sham ⫹ prompt laser group compared with
`each of the other 3 groups.
`Sample size was estimated to be 842 eyes (⬃701 study partic-
`ipants assuming 20% of study participants would have 2 study
`eyes) on the basis of an expected population difference in the letter
`score of 6.0 and standard deviation of the visual acuity letter score
`of 18, a correlation between baseline and 1-year scores of 0.48, a
`type 1 error rate of 0.016 (adjusted for multiple comparisons and
`alpha spending for interim data reviews), and a power of approx-
`imately 90%.
`The primary analysis included all randomized eyes and fol-
`lowed the intent-to-treat principle. Data were included in the
`1-year analysis when an examination was performed between 308
`and 420 days from randomization. When more than 1 visit oc-
`curred in this window, data from the visit closest to the 1-year
`target date were used. For eyes without 1-year data, the last-
`observation-carried forward method was used to impute data for
`the primary analysis. Similar results (data not shown) were pro-
`duced when analyses (1) used Rubin’s method14 to impute for
`missing data; (2) included only eyes with a completed 1-year
`
`Celltrion Exhibit 1018
`Page 3
`
`

`

`DRCR Network Writing Committee 䡠 Laser-Ranibizumab-Triamcinolone RCT for DME
`
`examination and used the last visual acuity before additional
`treatment for those who received a treatment other than the ran-
`domly assigned treatment before the 1-year examination (per-
`protocol analysis); (3) included adjustment for the following po-
`tential confounders in addition to baseline visual acuity: age,
`gender, race/ethnicity, baseline hemoglobin A1c, baseline OCT
`central subfield thickness, and prior panretinal scatter photocoag-
`ulation and prior DME treatment at baseline; (4) were performed
`with outlying values truncated to 3 standard deviations from the
`mean; and (5) used van der Waerden’s normal score transforma-
`tion on the visual acuity scores. For analyses other than the
`primary analysis, only data from completed visits were used with
`no imputation for missing data. For some results, medians and
`interquartile ranges have been reported instead of, or in addition to,
`means and standard deviations to describe the distribution of the
`data. Analyses of the number of study treatments received before
`the 1- and 2-year visits included only the eyes of participants
`completing the 1- and 2-year visits.
`Three pairwise comparisons were made for all analyses, except
`the ranibizumab groups were pooled for analysis of progression of
`diabetic retinopathy and all safety analyses. For all continuous
`outcomes, treatment group comparisons were made using analysis
`of covariance models with generalized estimating equations to
`account for correlated data from study participants with 2 study
`eyes. For binary outcomes, proportions similarly were compared
`between treatment groups using logistic regression models with
`generalized estimating equations. All analyses included adjustment
`for baseline visual acuity. In addition, models in which the central
`subfield thickness was the outcome included baseline central sub-
`field thickness as a covariate, and models with retinal volume as
`the outcome included both baseline central subfield thickness and
`retinal volume as covariates. Similar analyses were performed on
`2-year results. All P values are 2-sided. SAS version 9.1 (SAS Inc,
`Cary, NC) was used for all analyses.
`
`Results
`
`Between March of 2007 and December of 2008, 691 study par-
`ticipants (mean age 63⫾10 years; 44% women) were enrolled, 163
`(24%) with 2 study eyes. The mean baseline visual acuity letter
`score in study eyes was 63⫾12 (⬃20/63⫾2.4 lines), and the mean
`OCT central subfield retinal thickness was 405⫾134 ␮m. The 854
`study eyes were assigned to either sham ⫹ prompt laser (n⫽293),
`ranibizumab ⫹ prompt laser (n⫽187), ranibizumab ⫹ deferred
`laser (n⫽188), or triamcinolone ⫹ prompt laser (n⫽186). The
`baseline characteristics of the 4 groups were similar (Table 2,
`available at http://aaojournal.org).
`
`Follow-Up
`The follow-up status for all study participants (eyes) is shown in
`Figure 1 (available at http://aaojournal.org). Thirteen study partic-
`ipants (2%) died before the 1-year primary outcome visit and 15
`participants died subsequently of causes apparently unrelated to
`study treatment. For the remaining study participants, the 1-year
`primary outcome visit was completed for 94% to 96% of eyes in
`the 4 treatment groups. Those who completed the 1-year primary
`outcome visit completed 94% of the non-annual visits before 1
`year. Baseline visual acuity was similar in the 55 study eyes of the
`44 study participants who did not complete the 1-year primary
`outcome visit compared with the 799 eyes of the 647 study
`participants who completed the 1-year primary outcome visit (data
`not shown). The 2-year visit was completed for 484 eyes (57%),
`with 267 (31%) still pending, as of February 8, 2010.
`
`Treatments
`
`Sham Injections and Intravitreal Study Drug Injections. For
`each study participant, there were 13 possible sham or study drug
`injections during the first year of follow-up. The median (25th,
`75th percentile) number of sham injections before the 1-year
`primary outcome visit was 11 (8, 13) in the sham ⫹ prompt laser
`group (of note, this excludes 56 eyes among 163 participants with
`2 study eyes that were unmasked at baseline because the study
`participant’s other eye was in the ranibizumab ⫹ deferred laser
`group, precluding sham injections for the study eye assigned to
`sham ⫹ prompt laser). The median number of study drug injec-
`tions before the 1-year primary outcome visit was 8 (6, 10)
`ranibizumab injections (of 13 maximally possible injections) in the
`ranibizumab ⫹ prompt laser group, 9 (6, 11) ranibizumab injec-
`tions (of 13 maximally possible injections) in the ranibizumab ⫹
`deferred laser group, and 5 (3, 7) sham injections (of 9 maximally
`possible sham injections) and 3 (2, 4) triamcinolone injections (of
`4 maximally possible triamcinolone injections) for a total of 13
`maximally possible sham plus triamcinolone injections in the
`triamcinolone ⫹ prompt
`laser group (Fig 2, available at
`http://aaojournal.org).
`Retreatments Relative to ‘Success’ and ‘Failure’ Criteria. At
`the 16-week study visit, 47 (25%) of the 187 eyes in the ranibi-
`zumab ⫹ prompt laser group and 41 (22%) of the 188 eyes in the
`ranibizumab ⫹ deferred laser group met ‘success’ criteria (visual
`acuity letter score ⱖ84 [⬃ⱖ20/20] or OCT central subfield ⬍250
`␮m) and did not receive an injection. A total of 17 eyes (9%) in the
`ranibizumab ⫹ prompt laser group and 15 eyes (8%) in the
`ranibizumab ⫹ deferred laser group met ‘success’ criteria at 16
`weeks and did not receive an additional injection before the 1-year
`primary outcome visit. At the 1-year primary outcome visit, 89
`(32%) of the eyes in the sham ⫹ prompt laser group, 109 (64%) of
`the eyes in the ranibizumab ⫹ prompt laser group, 92 (52%) of the
`eyes in the ranibizumab ⫹ deferred laser group, and 98 (56%) of
`the eyes in the triamcinolone ⫹ prompt laser group met the
`‘success’ criteria, including 23 (8%), 23 (13%), 23 (13%), and 19
`(11%), respectively, with a visual acuity letter score ⱖ84 (⬃ⱖ20/
`20). ‘Failure’ criteria were met in 10 (4%), 3 (2%), 1 (1%), and 3
`(2%) of the eyes in these 4 groups, respectively, during the first
`year of follow-up. Sham or study drug injections were not required
`for eyes meeting ‘success’ or ‘failure’ criteria.
`Retreatments through Year 2. For the 218 study participants
`(58%) with 2 years of follow-up in the ranibizumab groups, there
`was a maximum of 25 possible ranibizumab injections. The me-
`dian (25th, 75th percentile) number of ranibizumab injections
`between the 1-year visit, inclusive, and before the 2-year visit were
`2 (0, 4) and 3 (1, 7) in the ranibizumab ⫹ prompt laser group and
`the ranibizumab ⫹ deferred laser group, respectively, for a total
`of 11 (7, 14) and 13 (8, 17) injections from baseline to the
`2-year visit. Only 32% of participants in the ranibizumab ⫹
`prompt laser group and 21% of participants in the ranibizumab
`⫹ deferred laser group had no ranibizumab injections between
`the 1- and 2-year visits. The 103 study participants (55%) with 2
`years of follow-up in the triamcinolone ⫹ prompt laser group
`received 1 (0, 2) triamcinolone injection between the 1-year visit,
`inclusive, and before the 2-year visit for a total of 4 (3, 5) from
`baseline to the 2-year visit of a total of 8 maximum possible
`injections.
`Focal/Grid Laser Treatments. The distribution of laser treat-
`ments before the 1- and 2-year visits are shown in Table 3
`(available at http://aaojournal.org). The median (25th, 75th per-
`centile) number of focal/grid photocoagulation treatments before
`the 1-year primary outcome visit was 3 (2, 3) in the sham ⫹
`prompt laser group, 2 (1, 3) in the ranibizumab ⫹ prompt laser
`group, and 2 (1, 3) in the triamcinolone ⫹ prompt laser group. In
`
`1067
`
`Celltrion Exhibit 1018
`Page 4
`
`

`

`Ophthalmology Volume 117, Number 6, June 2010
`
`Table 5. Change in Visual Acuity (Last Observation Carried Forward) from Baseline to 1 Year (Primary Outcome)*
`
`Change in visual acuity (letters)
`Mean ⫾ SD
`Median (25th, 75th
`percentile)
`Difference in mean change from
`sham ⫹ prompt laser (95%
`CI) [P value]†
`Distribution of change, No.
`(%)
`ⱖ15 letter improvement
`14–10 letter improvement
`9–5 letter improvement
`Same ⫾4 letters
`5–9 letters worse
`10–14 letters worse
`ⱖ15 letters worse
`Difference in proportion with
`ⱖ10 letter improvement
`from sham ⫹ prompt laser
`(95% CI)‡
`Relative risk (95% CI)
`[P value]§ for comparison
`with sham ⫹ prompt laser
`Difference in proportion with
`ⱖ10 letter worsening from
`sham ⫹ prompt laser (95%
`CI)‡
`Relative risk (95% CI)
`[P value]‡ for comparison
`with sham ⫹ prompt laser
`Difference in proportion with
`ⱖ15 letter improvement
`from sham ⫹ prompt laser
`(95% CI)‡
`Relative risk (95% CI)
`[P value]§ for comparison
`with sham ⫹ prompt laser
`Difference in proportion with
`ⱖ15 letter worsening from
`sham ⫹ prompt laser (95%
`CI)‡
`Relative risk (95% CI)
`[P value]§ for comparison
`with sham ⫹ prompt laser
`
`Sham ⴙ
`Prompt Laser
`N⫽293
`
`⫹3⫾13
`⫹5 (⫺2, ⫹10)
`
`Ranibizumab ⴙ
`Prompt Laser
`N⫽187
`
`⫹9⫾11
`⫹10 (⫹3, ⫹16)
`
`Ranibizumab ⴙ
`Deferred Laser
`N⫽188
`
`⫹9⫾12
`⫹9 (⫹5, ⫹15)
`
`Triamcinolone ⴙ
`Prompt Laser
`N⫽186
`
`⫹4⫾13
`⫹5 (⫺3, ⫹12)
`
`⫹5.8 (⫹3.2 to ⫹8.5)
`[P⬍0.001]
`
`⫹6.0 (⫹3.4 to ⫹8.6)
`[P⬍0.001]
`
`⫹1.1 (⫺1.5 to ⫹3.7)
`[P⫽0.31]
`
`43 (15%)
`38 (13%)
`67 (23%)
`86 (29%)
`20 (7%)
`16 (5%)
`23 (8%)
`
`57 (30%)
`38 (20%)
`34 (18%)
`38 (20%)
`14 (7%)
`3 (2%)
`3 (2%)
`⫹23% (⫹13% to ⫹34%)
`
`52 (28%)
`36 (19%)
`54 (29%)
`35 (19%)
`5 (3%)
`2 (1%)
`4 (2%)
`⫹19% (⫹9% to ⫹29%)
`
`39 (21%)
`22 (12%)
`32 (17%)
`54 (29%)
`12 (6%)
`12 (6%)
`15 (8%)
`⫹6% (⫺4% to ⫹16%)
`
`1.0
`
`1.0
`
`1.0
`
`1.0
`
`1.84 (1.40 to 2.42)
`[P⬍0.001]
`⫺10% (⫺16% to ⫺5%)
`
`1.68 (1.27 to 2.21)
`[P⬍0.001]
`⫺10% (⫺16% to ⫺4%)
`
`1.21 (0.88 to 1.66)
`[P⫽0.16]
`⫹1% (⫺7% to ⫹9%)
`
`0.24 (0.09 to 0.65)
`[P⬍0.001]
`⫹16% (⫹6% to ⫹26%)
`
`0.24 (0.08 to 0.68)
`[P⫽0.001]
`⫹13% (⫹4% to ⫹22%)
`
`1.08 (0.62 to 1.87)
`[P⫽0.75]
`⫹6% (⫺2% to ⫹15%)
`
`2.09 (1.35 to 3.22)
`[P⬍0.001]
`⫺6% (⫺11% to ⫺2%)
`
`1.89 (1.25 to 2.87)
`[P⬍0.001]
`⫺6% (⫺10% to ⫺1%)
`
`1.43 (0.90 to 2.29)
`[P⫽0.07]
`0 (⫺6% to ⫹6%)
`
`0.21 (0.05 to 0.87)
`[P⫽0.009]
`
`0.28 (0.08 to 0.97)
`[P⫽0.01]
`
`1.02 (0.47 to 2.20)
`[P⫽0.95]
`
`CI ⫽ confidence interval; SD ⫽ standard deviation.
`*Visits occurring between 308 and 420 days (between 44 and 60 wks) from randomization were included as 1-yr visits. When ⬎ 1 visit occurred in this
`window, data from the visit closest to the 1-yr target date were used. For other eyes without any 1-yr data (19 eyes in the sham ⫹ prompt laser group,
`16 eyes in the ranibizumab ⫹ prompt laser group, 10 eyes in the ranibizumab ⫹ deferred laser group, and 10 eyes in the triamcinolone ⫹ prompt laser
`group), the last observation carried forward method was used to impute data for the primary analysis.
`†Analysis of covariance adjusted for baseline visual acuity and correlation between 2 study eyes. Confidence intervals are adjusted for multiple comparisons.
`‡Adjusted for correlation between 2 study eyes. Confidence intervals are adjusted for multiple comparisons.
`§Logistic regression adjusted for correlation between 2 study eyes. Confidence intervals are adjusted for multiple comparisons.
`
`the ranibizumab ⫹ prompt laser group, after baseline and before the
`1-year primary outcome visit, 53 (31%) study eyes received no
`additional focal/grid laser treatments, 54 (32%) received only 1 ad-
`ditional focal/grid laser treatment, 46 (27%) received only 2 additional
`focal/grid laser treatments, and 18 (11%) received 3 additional focal/
`grid laser treatments. Focal/grid laser treatment was not permitted in
`the ranibizumab ⫹ deferred laser group until the 24-week study visit;
`from the 24-week study visit and before the 1-year primary outcome
`
`visit, 128 (72%) of these study eyes received no focal/grid laser
`treatment, 35 (20%) received only 1 focal/grid laser treatment,
`and 15 (8%) received 2 focal/grid laser treatments. Forty-seven
`percent of the sham ⫹ prompt laser group, 57% of the ranibi-
`zumab ⫹ prompt
`laser group, 72% of the ranibizumab ⫹
`deferred laser group, and 46% of the triamcinolone ⫹ prompt
`laser group received no focal/grid laser treatments between the
`1- and 2-year visits.
`
`1068
`
`Celltrion Exhibit 1018
`Page 5
`
`

`

`DRCR Network Writing Committee 䡠 Laser-Ranibizumab-Triamcinolone RCT for DME
`
`Figure 3. Mean change in visual acuity at follow-up visits. Values that were ⫾30 letters were assigned a value of 30. P values for difference in mean
`change in visual acuity from sham ⫹ prompt laser at 52 weeks: ranibizumab ⫹ prompt laser ⬍0.001, ranibizumab ⫹ deferred laser ⬍0.001, and
`triamcinolone ⫹ prompt laser groups ⫽ 0.31. Each visit week includes visits that are ⫾14 days, except the 52-week visit, which includes visits that
`occur between 308 and 420 days (between 44 and 60 weeks) from randomization, and the 104-week visit, which includes visits that occur between
`616 and 840 days (between 88 and 120 weeks) from randomization.
`
`Alternative Treatments. Some eyes
`in the study were
`switched from the randomly assigned treatment to an alternative
`treatment during the first 2 years of follow-up because “failure” or
`“futility” criteria were met or the treating investigator determined
`deviating from the protocol would be in the best interest of the
`study participant as a patient. In the sha