Evaluation of Injection Frequency and
`Visual Acuity Outcomes for Ranibizumab
`Monotherapy in Exudative Age-related
`Macular Degeneration
`
`Hajir Dadgostar, MD, PhD, Alexandre A. C. M. Ventura, MD, Jeffrey Y. Chung, MD, Sumit Sharma, BS,
`Peter K. Kaiser, MD
`
`Objective: To evaluate the visual outcomes for intravitreal ranibizumab administered on an as-needed basis
`for exudative age-related macular degeneration (AMD) and to investigate the relationship between injection
`frequency and visual outcome in this setting.
`Design: Retrospective, interventional case series.
`Participants: A total of 131 eyes with treatment-naïve, exudative AMD undergoing ranibizumab
`monotherapy.
`Methods:
`Intravitreal ranibizumab was administered on an as-needed basis guided by clinical examination
`and optical coherence tomography (OCT). The OCT scans were evaluated by the treating physicians for the
`presence of intraretinal fluid, subretinal fluid, intraretinal cysts, or increasing pigment epithelial detachment size.
`Clinical data including visual acuity (VA), choroidal neovascularization lesion morphology, and treatment course
`were collected retrospectively for analysis.
`Main Outcome Measures: Mean change in best-corrected Snellen VA.
`Results: The mean age was 81.3 years, mean follow-up was 12⫾4.3 months (minimum 6 months, median
`12 months), and mean number of injections was 5.2⫾2.8. Mean baseline Snellen VA for the entire population was
`20/110 and significantly improved at 6 months (20/80; P ⫽ 0.0002) and at last follow-up (20/90; P ⫽ 0.0066). At
`6 months, 31% of eyes had gained at least 3 lines of VA and 90.5% had avoided loss of 3 lines. On average, it
`took 3.0 injections and 3.5 months to achieve a “dry” or “flat” macula on OCT after initiating treatment. Resolution
`of intra- and subretinal fluid on OCT did not correlate with the degree of vision improvement. Eyes receiving more
`frequent injections (defined as ⬍2 months mean inter-injection interval) gained more vision (⫹2.3 lines at 6
`months) than eyes receiving injections less frequently (⫹0.46 lines at 6 months; P ⫽ 0.012). At 6 months, 3.1%
`of those in the more frequent injection group lost ⬎3 lines of vision compared with 15.9% in the ⬎2 months
`interval group (P ⫽ 0.011).
`Conclusions:
`In a population receiving as-needed injections of ranibizumab for exudative AMD, visual
`improvement was related to the frequency of injections received but not to the resolution of fluid by OCT.
`Treatment with ranibizumab on a strictly as-needed basis may result in undertreatment and significantly less
`visual gain.
`Financial Disclosure(s): Proprietary or commercial disclosure may be found after the references.
`Ophthalmology 2009;116:1740 –1747 © 2009 by the American Academy of Ophthalmology.
`
`Age-related macular degeneration (AMD) is the leading
`cause of blindness among the aging population.1 The intro-
`duction of molecular inhibitors of vascular endothelial
`growth factor (VEGF), such as pegaptanib, ranibizumab,
`and bevacizumab, as approved and off-label therapy for
`choroidal neovascularization (CNV) due to exudative AMD
`has revolutionized our management of these patients.2 The
`phase III pivotal clinical trials, Minimally Classic/Occult
`Trial of the Anti-VEGF Antibody Ranibizumab in the
`Treatment of Neovascular AMD (MARINA) and Anti-
`VEGF Antibody for the Treatment of Predominantly Clas-
`sic Choroidal Neovascularization in AMD (ANCHOR),
`have demonstrated the efficacy and safety of monthly
`
`ranibizumab injections (Lucentis, Genentech, South San
`Francisco, CA) for the preservation and improvement of
`visual acuity (VA) in patients with CNV due to exudative
`AMD.3,4
`These key trials did not incorporate a treatment end point
`and continued injections irrespective of clinical course.
`Some emerging evidence has supported the adoption of a
`variable VEGF inhibitor dosing strategy guided by serial
`diagnostic reevaluation by optical coherence tomography
`(OCT).5 The Prospective OCT Imaging of Patients with
`Neovascular AMD Treated with Intra-ocular Lucentis
`(PrONTO) study was a prospective investigator-sponsored
`trial in 40 eyes that demonstrated 3 lines or more VA gain
`
`1740
`
`© 2009 by the American Academy of Ophthalmology
`Published by Elsevier Inc.
`
`ISSN 0161-6420/09/$–see front matter
`doi:10.1016/j.ophtha.2009.05.033
`
`Mylan Exhibit 1094
`Mylan v. Regeneron, IPR2021-00880
`Page 1
`
`

`

`Dadgostar et al
`
`䡠 Ranibizumab Injection Frequency and Visual Acuity
`
`in 35% of patients and avoidance of 3 lines VA loss in 95%
`of eyes receiving 3 monthly ranibizumab injections fol-
`lowed by as-needed treatment based on clinical examination
`and serial OCT monitoring.5 This as-needed dosing scheme
`based on OCT has largely been adopted without any clinical
`trial evidence. Ongoing studies are currently evaluating the
`effectiveness of as-needed dosing regimens compared with
`monthly dosing, different monitoring strategies, and com-
`bination therapies to determine the optimal niche for this
`new class of drugs in AMD management (Martin DF. The
`Comparison of Age-Related Macular Degeneration Treat-
`ments Trials [CATT]. Presented at: Annual Advanced Vit-
`reoretinal Techniques and Technology Conference, Septem-
`ber 8, 2007, Chicago, IL).
`The goal of the present study is to examine the effect of
`intravitreal ranibizumab monotherapy administered on an
`as-needed basis guided by OCT monitoring in treatment-
`naïve patients and to examine the relationship between
`dosing frequency and VA response to treatment in these
`patients.
`
`Materials and Methods
`
`After Cleveland Clinic Institutional Review Board approval, the
`charts of all patients with CNV due to exudative AMD receiving
`their first ranibizumab injection at the Cole Eye Institute (Cleve-
`land, OH) between April 2006 and December 2007 were retro-
`spectively reviewed. Patients were included in the study if they
`were older than 50 years of age, were treatment naïve and receiv-
`ing their first ranibizumab injection for exudative AMD, and had
`at least 6 months of follow-up. Patients were excluded if there was
`documented evidence of diabetic retinopathy, vascular occlusion,
`epiretinal membrane, active uveitis, or any prior treatment for
`exudative AMD in the study eye before the initiation of ranibi-
`zumab monotherapy. In addition, patients with coexisting diag-
`noses that may lead to CNV, including 1 patient with lesions
`suggestive of presumed ocular histoplasmosis syndrome, 1 patient
`with myopic degeneration, 2 patients with vitelliform lesions, 1
`patient with familial pattern dystrophy, and 1 patient with pre-
`sumed idiopathic CNV, were excluded.
`All eyes were treated with intravitreal injections of 0.5 mg
`ranibizumab per published guidelines. Injections were adminis-
`tered once at baseline and then on an as-needed basis at the
`discretion of the treating physician according to serial clinical
`evaluation and OCT monitoring using the Stratus OCT (software
`version 4.1, Carl Zeiss Meditec, Inc., Dublin, CA). All scans were
`obtained through a dilated pupil by experienced certified OCT
`photographers using Digital OCT Reading Center (DOCTR) Pro-
`tocols consisting of a fast macular thickness map, which acquires
`six 6-mm radial lines consisting of 128 A-scans per line in 1.92
`seconds of scanning; 3- and 6-mm horizontal and vertical linear
`cross-hair scans centered through the fovea as determined by
`evaluation of the red-free image on the computer monitor of the
`OCT scanner; and a custom posterior pole scan consisting of a
`7-mm custom line scan that extends 5 degrees below horizontal
`from the temporal edge of the optic nerve toward the fovea. This
`custom scan pattern is based on a line extending from the center
`temporal edge of the optic nerve 5 degrees inferior-to-horizontal
`temporally. The treating physician evaluated the OCT images, and
`treatment was administered if there was evidence of any increased
`retinal thickening (i.e., intraretinal fluid), subretinal fluid (SRF),
`intraretinal cysts, or increasing pigment epithelial detachment
`(PED). Treatment was also delivered if there was clinical evidence
`
`of new or increasing subretinal or intraretinal hemorrhage on
`clinical examination.
`Charts were reviewed for baseline demographic data, coexist-
`ing diagnoses, lens status, number of injections, and best-corrected
`Snellen VA at baseline, 1 month (⫾2 weeks), 2 months, 3 months,
`and 6 months (⫾4 weeks) after initiation of treatment and at the
`last documented follow-up visit. Snellen VA was converted to the
`logarithm of the minimal angle of resolution (logMAR) scale for
`statistical analyses based on formulas set forth previously.6 Base-
`line fluorescein angiograms (FAs) were graded independently by 2
`retina specialists (AV and JC) to determine CNV lesion composi-
`tion, location, and size in disc areas using OIS WinStation Soft-
`ware (Medivision-OIS, Inc., Sacramento, CA). Baseline OCT
`scans were analyzed by 2 certified DOCTR readers (HD and SS)
`using DOCTR protocols for manual measurements of central ret-
`inal thickness, CNV thickness, SRF thickness, and PED height
`using Stratus OCT Reading Software version 4.1 (Carl Zeiss
`Meditec, Inc.).
`Mean best-corrected VA compared with baseline was calcu-
`lated using paired t test analysis. Unpaired t test analyses were
`performed for comparisons of mean VA change among groups
`based on mean injection frequency and for comparisons of CNV
`size and OCT lesion thickness among subgroups. Injection
`number was compared among anatomic treatment response
`groups using 1-way analysis of variance (ANOVA). Pairwise
`correlations were performed using Spearman’s coefficient.
`Analyses of proportions of eyes with vision gain or loss were
`performed using the likelihood-ratio chi-square statistic. All
`statistical analyses were performed using JMP Software Ver-
`sion 7.0 (SAS Inc., Cary, NC).
`
`Results
`
`A total of 446 patients were reviewed, and 131 treatment-naïve
`eyes in 124 patients met the inclusion criteria and were included in
`the analysis. Demographic and clinical data for these patients are
`summarized in Table 1. The mean age was 81.3⫾8.3 years, and the
`gender distribution was 64.5% female. Baseline FAs were avail-
`able for 104 eyes, and these revealed a mix of baseline lesion types
`(16.3% predominantly classic CNV, 16.3% minimally classic
`CNV, 66.3% occult with no classic CNV) that were mostly sub-
`foveal or juxtafoveal in location (75% subfoveal, 20.2% juxtafo-
`veal, and 4.8% extrafoveal). The mean baseline center subfield
`
`Table 1. Patient, Lesion, and Treatment Characteristics
`
`Age
`Gender
`Diabetes mellitus
`Glaucoma
`History of uveitis
`Follow-up (mos)
`Injections (total)
`Time to dry (mos)*
`Injections to dry*
`Baseline FA: lesion type
`
`Baseline FA: lesion loc.
`
`Median 83 yrs, mean 81.3⫾8.3 yrs
`44/124 (35.5%) male, 80/124 (64.5%) female
`15/124 (12.1%)
`16/124 (12.9%)
`3/124 (2.4%)
`Median 12, mean 12.0⫾4.3
`Mean 5.2⫾2.8
`Mean 3.5⫾2.6
`Mean 3.0⫾2.0
`17/104 (16.3%) classic, 17/104 (16.3%) min.
`classic, 69/104 (66.3%) occult
`78/104 (75%) subfoveal, 21/104 (20.2%)
`juxtafoveal, 5/104 (4.8%) extrafoveal
`
`FA ⫽ fluorescein angiogram; OCT ⫽ optical coherence tomography.
`Mean data are presented as mean ⫾ standard deviation.
`*Subgroups demonstrating resolution of subretinal and intraretinal fluid by
`OCT and clinical examination (groups 1 and 2; n ⫽ 98).
`
`1741
`
`Mylan Exhibit 1094
`Mylan v. Regeneron, IPR2021-00880
`Page 2
`
`

`

`Ophthalmology Volume 116, Number 9, September 2009
`
`Table 2. Visual Acuity Change and Anatomic Subgroups
`
`Anatomic Subgroups*
`
`Total
`
`Group 1
`(No Recurrence)
`
`Group 2
`(Recurrent Fluid)
`
`Group 3
`(Always Wet)
`
`No. (%)
`Mos F/U mean (SD)
`Mean VA baseline
`Mean VA – month 1 (P vs. baseline)†
`Mean VA – month 2 (P vs. baseline)
`Mean VA – month 3 (P vs. baseline)
`Mean VA – month 6 (P vs. baseline)
`Mean VA – last F/U (P vs. baseline)
`Month 3
`ⱖ3 lines gained (%)
`⬍3 lines lost (%)
`Month 6
`ⱖ3 lines gained (%)
`⬍3 lines lost (%)
`Last F/U
`ⱖ3 lines gained (%)
`⬍3 lines lost (%)
`
`131/131 (100)
`12.0 (4.3)
`0.75 [⬃20/110]
`0.65 [⬃20/90] (0.0013)
`0.60 [20/80] (⬍0.0001)
`0.61 [20/80] (⬍0.0001)
`0.60 [⬃20/80] (0.0002)
`0.65 [⬃20/90] (0.0066)
`
`36/131 (27.5)
`10.4 (3.7)
`0.79 [⬃20/125]
`0.71 [⬃20/100] (0.075)
`0.64 [⬃20/90] (0.0091)
`0.71 [⬃20/100] (0.10)
`0.73 [⬃20/110] (0.26)
`0.77 [⬃20/115] (0.42)
`
`62/131 (47.3)
`14.2 (3.9)
`0.63 [⬃20/85]
`0.56 [⬃20/70] (0.057)
`0.51 [⬃20/63] (0.0012)
`0.51 [⬃20/63] (0.0012)
`0.46 [⬃20/60] (⬍0.0001)
`0.54 [20/70] (0.05)
`
`33/131 (25.2)
`9.7 (3.5)
`1.02 [⬃20/200]
`0.84 [⬃20/140] (0.0093)
`0.82 [⬃20/125] (0.0095)
`0.74 [⬃20/110] (0.0007)
`0.79 [⬃20/125] (0.06)
`0.76 [20/114] (0.012)
`
`33/120 (27.5)
`110/120 (91.7)
`
`39/126 (31.0)
`114/126 (90.5)
`
`39/130 (30.0)
`109/130 (83.8)
`
`8/33 (24.2)
`30/33 (90.9)
`
`11/34 (32.4)
`29/34 (85.3)
`
`8/36 (22.2)
`29/36 (80.6)
`
`16/57 (28.1)
`51/57 (89.5)
`
`19/60 (31.7)
`56/60 (93.3)
`
`20/61 (32.8)
`51/61 (83.6)
`
`9/30 (30.0)
`29/30 (96.7)
`
`9/32 (28.1)
`29/32 (90.6)
`
`11/33 (33.3)
`29/33 (87.9)
`
`VA ⫽ visual acuity (mean logMAR format with Snellen conversion in brackets); SD ⫽ standard deviation; F/U ⫽ follow-up.
`*Subgroups defined by anatomic response to treatment. Group 1 ⫽ resolution of subretinal/intraretinal fluid by OCT and clinical examination without
`recurrence, no further injections after “dry.” Group 2 ⫽ resolution of fluid at some time with at least 1 recurrence based on OCT and clinical examination.
`Group 3 ⫽ persistent subretinal or intraretinal fluid throughout the follow-up period.
`†Paired t test analyses restricted to datasets with observations at all timepoints (total n ⫽ 99; group 1 n ⫽ 26; group 2 n ⫽ 53; group 3 n ⫽ 20).
`
`thickness by OCT was 299.2 ␮m. Detectable SRF was present in
`75.7% of eyes at baseline, and a PED was present in 68.9%.
`Both mean and median follow-ups from the time of the first
`injection were 12 months. The mean total number of injections was
`5.2⫾2.8 over this follow-up period. For those eyes demonstrating
`complete resolution of intraretinal fluid and SRF on follow-up
`OCT (i.e., “dry” or “flat” OCT), the mean number of injections to
`achieve this anatomic end point was 3.0⫾2.0 injections over
`3.5⫾2.6 months from the time of the initial injection.
`As shown in Table 2, the mean baseline VA was 20/110 and
`there was a statistically significant improvement in vision to 20/90
`as early as month 1 (i.e., after a single injection; paired t test, P ⫽
`0.0013). At month 6, average VA was 20/80 (P ⫽ 0.0002) and
`mean VA at last follow-up was approximately 20/90 (P ⫽ 0.0066).
`The mean change in Snellen VA from baseline was ⫹1 line, ⫹1.5
`lines, ⫹1.4 lines, ⫹1.5 lines, and ⫹ 1 line at months 1, 2, 3, and
`6, and last follow-up, respectively. At 6 months, 31% of eyes had
`gained at least 3 lines of VA, 57.1% had gained more than 0 letters,
`and 90.5% had avoided the loss of 3 lines or more of VA. The
`proportion of eyes with 3 lines or more gained remained 30% at
`last follow-up; however,
`the proportion gaining any VA was
`49.2% and only 83.8% avoided loss of 3 or more lines VA at last
`follow-up.
`On the basis of the anatomic response to treatment (as deter-
`mined by the presence of any retinal fluid on follow-up OCT), we
`defined 3 subgroups (Table 2). Group 1 (n ⫽ 36) included those
`eyes that showed completely resolved intraretinal fluid and SRF,
`that showed no evidence of recurrence clinically or by OCT on
`serial reexaminations during the study follow-up, and that required
`no further treatment. Group 2 (n ⫽ 62) included those eyes that
`demonstrated resolution of fluid as described previously, but then
`had at least 1 recurrence of either subretinal or intraretinal fluid on
`repeat examination during follow-up that required reinitiation of
`injections after they were initially stopped. Group 3 (n ⫽ 33)
`included those eyes that had persistent leakage by OCT throughout
`the follow-up period, and injections were never stopped. These
`groups were defined for demonstrative purposes within the context
`
`1742
`
`of this study, and it is understood that given a sufficiently long
`follow-up the majority of eyes in group 1 would likely cross over
`into group 2 as defined previously.
`Although all groups demonstrated some improvement in VA
`with treatment, there was a clear trend toward greater improvement
`in groups 2 and 3 (Fig 1). Although group 1 appeared to exhibit
`moderate VA improvement within the first few months, the vision
`gradually drifted toward baseline VA (20/125) by the last follow-
`up (20/115, P ⫽ 0.42). At last follow-up, 22.2% of eyes in
`group 1 had gained at least 3 lines of VA, compared with 32.8%
`and 33.3% in groups 2 and 3, respectively. Although there was a
`trend toward greater VA improvement in groups 2 and 3 compared
`
`Figure 1. Visual acuity change by anatomic response group. Mean VA
`change from baseline was plotted for each group at the indicated time
`intervals and at last follow-up. Error bars indicate standard error. VA ⫽
`visual acuity; logMAR ⫽ logarithm of the minimal angle of resolution;
`F/U ⫽ follow-up.
`
`Mylan Exhibit 1094
`Mylan v. Regeneron, IPR2021-00880
`Page 3
`
`

`

`Dadgostar et al
`
`䡠 Ranibizumab Injection Frequency and Visual Acuity
`
`Figure 2. Injection number by anatomic response group. Scatter diagrams depict total number of injections (A) and number of injections normalized by
`follow-up period to approximate injections per year (B) for anatomic response groups 1, 2, and 3 (defined above). Diamonds indicate means and 95% CIs
`for each group. Outer lines indicate standard deviation. All means were compared by 1-way ANOVA.
`
`with group 1, the differences in VA change at month 3, month 6,
`and last follow-up were not significantly different between the 3
`groups (ANOVA, all P⬎0.4; pairwise t tests, all P⬎0.2).
`When the mean number of injections was compared among
`groups 1 (3.3⫾2.0), 2 (5.9⫾2.6), and 3 (6.1⫾3.0), there were
`significantly more injections in groups 2 and 3 (Fig 2; ANOVA,
`P⬍0.0001). A normalized measure representing the average num-
`ber of injections per year also was calculated for every eye in each
`group. By using this measurement, the normalized injections for
`groups 1 (4.3⫾2.7), 2 (5.2⫾2.1), and 3 (7.6⫾3.0) also were
`significantly different (ANOVA, P⬍0.0001). Normalized injec-
`tions per year then were compared against VA change at last
`follow-up, and bivariate analysis revealed a weak association with
`a trend toward greater VA gain with a larger number of injections
`per year (Spearman r ⫽ 0.21, P ⫽ 0.014).
`An approximation of injection frequency was obtained by
`dividing the total follow-up duration by the total number of injec-
`tions to determine the relationship between injection frequency
`and VA change. The study population was then divided into 2
`groups on the basis of injection frequency: those who received
`repeat injections at a mean interval of ⱕ2 months throughout the
`follow-up period (⬍2 months; n ⫽ 65) and those with an average
`injection interval of ⬎2 months (⬎2 months; n ⫽ 66). In the ⬍2
`months and ⬎2 months groups, the percentages of female patients
`were 60% and 69.7%, respectively (P ⫽ 0.24), and the mean ages
`were 81⫾8.4 years and 81.5⫾8.0 years, respectively (P ⫽ 0.74).
`The mean baseline VA (logMAR) was 0.70⫾0.50 for the ⬍2
`months group and 0.84⫾0.57 for the ⬎2 months group (P ⫽ 0.15).
`Baseline OCT characteristics were compared among the 103 eyes
`for which baseline OCTs were available. Baseline central retinal
`thickness measurements for the ⬍2 months group (n ⫽ 52) and ⬎2
`months group (n ⫽ 51) were 261 ␮m and 288 ␮m, respectively
`(P ⫽ 0.38), and there were no significant differences in the
`distribution of specific morphologic features on baseline OCT:
`SRF present in 78.9% versus 72.6%, respectively (P ⫽ 0.46);
`intraretinal fluid present in 90.4% versus 88.2%, respectively (P ⫽
`0.72); lipid exudates present in 61.5% versus 54.9%, respectively
`(P ⫽ 0.49); and PED present in 73.1% versus 64.7%, respectively
`(P ⫽ 0.36). Also, for those lesions with measurable CNV, SRF, or
`PED on baseline OCT, there was no statistically significant dif-
`ference between the ⬍2 months group and the ⬎2 months group
`in mean CNV thickness (P ⫽ 0.91), mean SRF thickness (P ⫽
`0.3), or mean PED height (P ⫽ 0.14). Baseline angiographic
`
`characteristics were compared among the 104 eyes for which
`baseline FAs were available. Mean CNV lesion sizes in the ⬍2
`months group (n ⫽ 53) and ⬎2 months group (n ⫽ 51) were 5.5 and
`5.6 disc areas, respectively (P ⫽ 0.95). The percentages of pre-
`dominantly classic, minimally classic, occult, and indeterminate
`CNV types were 11.3%, 18.9%, 67.9%, and 1.9%, respectively, for
`the ⬍2 months group and 21.6%, 13.7%, 64.7%, and 0%, respec-
`tively, for the ⬎2 months group, (P ⫽ 0.32). Lesion locations were
`subfoveal,
`juxtafoveal, and extrafoveal
`in 79.3%, 17.0%, and
`3.8%, respectively, for the ⬍2 months group and 70.6%, 23.5%,
`and 5.9%, respectively, for the ⬎2 months group (P ⫽ 0.59). Thus,
`we were unable to identify any baseline feature that differed
`significantly between these 2 groups.
`As shown in Figure 3, those who received repeat injections at
`a mean interval of 2 months or less throughout the follow-up
`period had a greater improvement in VA at month 6 (mean change
`in vision ⫹2.3 lines, 95% confidence interval [CI] ⫹0.14 to ⫹0.32
`logMAR) and at last follow-up (mean change in vision ⫹2.1 lines,
`95% CI ⫹0.11 to ⫹0.31 logMAR) compared with those with ⬎2
`months average injection interval at 6 months (mean change in
`vision ⫹0.46 lines, 95% CI ⫺0.087 to ⫹0.18 logMAR; P ⫽
`0.012) and at last follow-up (mean change in vision ⫺0.22 lines,
`95% CI ⫺0.14 to ⫹0.095 logMAR; P ⫽ 0.0015). Similarly, there
`was a trend toward a higher percentage of eyes gaining at least 3
`lines of VA in the ⬍2 months injection interval group compared
`with the ⬎2 months group (37.5% vs. 24.2% at 6 months follow-
`up, P ⫽ 0.11), but this difference did not reach statistical signif-
`icance. The difference in percentage of eyes with 3 lines of VA
`loss, however, was statistically significant (Fig 4) with less loss in
`the ⬍2 months group (3.1% at 6 months, 7.8% at last follow-up)
`compared with the ⬎2 months group (15.9% at 6 months, P ⫽
`0.011; 24.2% at last follow-up, P ⫽ 0.0093).
`
`Discussion
`
`Determining the optimal dosing schedule for anti-VEGF
`therapy for exudative AMD remains a challenge because
`neither the effect of the drugs in human eyes nor the nature
`of the disease itself is fully understood. On the basis of the
`MARINA and ANCHOR trial data, the best reported effi-
`cacy, in terms of VA preservation and gain, is seen with
`
`1743
`
`Mylan Exhibit 1094
`Mylan v. Regeneron, IPR2021-00880
`Page 4
`
`

`

`Ophthalmology Volume 116, Number 9, September 2009
`
`Figure 3. Greater mean VA improvement with higher injection frequency. Scatter diagrams depict mean VA change (logMAR) from baseline at 6
`months (A) and at last follow-up (B) for eyes receiving injections at an average interval of ⱕ2 months versus an average interval of ⬎2 months over the
`follow-up period. Means were compared by unpaired t test. Inner error lines represent mean and standard error, and outer error lines represent standard
`deviation. VA ⫽ visual acuity; logMAR ⫽ logarithm of the minimal angle of resolution.
`
`mandated monthly dosing.3,4 However, these trials did not
`follow a clinical end point for treatment cessation, and such
`a regimen raises concerns for overtreatment. Because
`VEGF is a survival factor, long-term VEGF suppression
`may lead to ocular problems that we have not yet iden-
`tified. Although reported risks of injection-related ad-
`verse events such as retinal detachment or endophthalmi-
`tis are low,3,4,7,8 they are not zero, and the results can be
`devastating. Minimizing the number of injections would
`decrease this risk. Finally, overuse of such treatments is
`certain to present an unnecessarily heavy financial burden
`on society given the cost of ranibizumab.9,10
`
`Thus, for multiple reasons, it is desirable to find a dosing
`regimen that is maximally effective at preserving or improv-
`ing VA while at the same time is intended to avoid treat-
`ment when it is not needed. The Phase IIIb, Multi-center,
`Randomized, Double-masked, Sham Injection-controlled
`Study of the Efficacy and Safety of Ranibizumab (PIER)
`assessed the efficacy of fixed quarterly dosing by evaluating
`a regimen of 3 monthly injections on enrollment, followed
`by mandated injections every 3 months thereafter for a total
`of 2 years.11 In the ranibizumab groups, there was an initial
`mean improvement in VA by approximately 1 line after the
`loading phase of 3 monthly injections. However, at 12
`
`Figure 4. Moderate vision gain and loss by injection interval. Percentage of eyes in each group with at least 3 lines VA gain (A) or ⬎3 lines VA loss
`(B) at 6 months and at last follow-up for each injection frequency group was compared by likelihood-ratio chi-square test. VA ⫽ visual acuity; F/U ⫽
`follow-up.
`
`1744
`
`Mylan Exhibit 1094
`Mylan v. Regeneron, IPR2021-00880
`Page 5
`
`

`

`Dadgostar et al
`
`䡠 Ranibizumab Injection Frequency and Visual Acuity
`
`months the ranibizumab groups had essentially declined
`back to baseline VA, whereas sham-treated subjects lost
`approximately 16 letters (P⬍0.0001). Although clearly su-
`perior to placebo, the VA results with ranibizumab in this
`trial were markedly less impressive than in the MARINA
`and ANCHOR data. It
`is difficult
`to draw quantitative
`conclusions from such a comparison because PIER,
`MARINA, and ANCHOR were separate trials enrolling
`distinct populations. A direct comparison of these treat-
`ment schemes was performed in the Efficacy and Safety
`of Ranibizumab in Patients with Subfoveal Choroidal Neo-
`vascularization Secondary to Age-Related Macular Degen-
`eration (EXCITE) trial, a Phase IIIb multicenter, random-
`ized, active-controlled, double-masked trial of all CNV
`types that compared monthly injections with receiving 3
`monthly injections followed by quarterly injections. In this
`study, monthly injections of 0.3 mg ranibizumab (n ⫽ 101)
`resulted in a greater mean VA improvement from baseline
`at 12 months (⫹8.3 letters) compared with quarterly injec-
`tions of either 0.3 mg (n ⫽ 104) or 0.5 mg (n ⫽ 88)
`ranibizumab (⫹4.9 and ⫹3.8 letters, respectively) (Eter N.
`Ranibizumab for neovascular ARMD: results of the EXCITE
`and SUSTAIN studies. Presented at: German Retina Society,
`June 19–22, 2008, Wurzburg, Germany).
`Thus, monthly injections seem better than quarterly in-
`jections, but we still do not know how as-needed dosing fits.
`The investigator-sponsored PrONTO study featured serial
`OCT monitoring as a tool to guide repeat injections on a
`variable as-needed basis after 3 monthly doses.5 The visual
`results at 12 months were comparable to what has been
`reported with monthly injections, but these results were
`achieved with an average of 5.6 injections over 12 months.
`Similar results were recently reported in an interim analysis
`of data from the Study of Ranibizumab in Patients With
`Subfoveal Choroidal Neovascularization Secondary to Age-
`Related Macular Degeneration (SUSTAIN) study (Holz FG.
`Flexibly dosed ranibizumab in patients with neovascular
`AMD: twelve-month interim results of the SUSTAIN Trial.
`Presented at: American Academy of Ophthalmology, An-
`nual Meeting, November 8 –11, 2008, Atlanta, GA). In this
`phase IIIb, multicenter, open-label, 12-month study of
`ranibizumab for exudative AMD, a total of 531 eyes, some
`of which were treatment-naïve and some were ANCHOR
`completers, were treated with either 0.3 mg or 0.5 mg
`ranibizumab for 3 monthly loading doses and then reinjected as
`needed on the basis of protocol criteria of ⬎100 ␮m increased
`thickening on OCT or ⬎5 letters VA loss. An interim report
`of safety and efficacy data from 69 treatment-naïve eyes that
`had completed 12 months of follow-up reported that the
`mean number of injections over 12 months was 5.3 (range
`2–10). The maximum improvement in mean VA occurred at
`month 3 (⫹9.2 letters), and by month 12 the mean VA
`change from baseline was ⫹6.7 letters.
`In the present study, we retrospectively examined VA
`change in 131 eyes with treatment-naïve CNV due to exu-
`dative AMD treated with ranibizumab monotherapy once at
`baseline and then on an as-needed basis with OCT guidance
`at a single center with an average of 12 months follow-up.
`After a mean of 5.2 injections, there was a significant
`improvement in vision at last follow-up (⬃1 line gain), and
`
`the percentage of eyes with at least 3 lines gained (29.8%)
`was similar to data in the MARINA trial. The majority of
`eyes avoided moderate vision loss, although the rate in this
`study (84%) was somewhat lower than the rates reported
`with monthly injections (⬃95% at 12 months in the
`MARINA and ANCHOR trials). In a recent prospective
`study investigating a similar regimen of 1 injection at base-
`line followed by repeat injections as needed, Rothenbuehler
`et al12 reported similar overall visual outcomes in their
`population of 138 patients. After a mean of 5.6 injections in
`the first year and 4.3 injections in the second year, they
`observed a mean VA gain of 6.3 letters at 24 months with
`30% gaining 3 lines or more and 85% avoiding 3 lines of
`loss. In this study, at 12 months, approximately 10% of eyes
`continued to exhibit SRF or cysts on OCT; however, sub-
`group analyses relating vision change to anatomic change
`and injection number were not reported.
`Our study population was subdivided on the basis of
`anatomic response to treatment to distinguish those lesions
`that had active leakage resolved (as determined by OCT and
`clinical examination), and who received no further injec-
`tions during follow-up, from those lesions that had recurrent
`fluid on follow-up and those in which intraretinal or SRF
`never resolved throughout the follow-up period. The group
`without recurrences (also by definition the group receiving
`the fewest injections; Fig 1) exhibited least VA improve-
`ment compared with the other 2 groups. This suggests that
`anatomic outcomes and functional outcomes may not coin-
`cide, and what some may consider anatomic “failures” (e.g.,
`refractory or recurrent fluid) may actually be associated
`with favorable visual outcomes if continued to be treated
`appropriately.
`One variable that may affect visual outcomes among the
`anatomic response groups is the number of injections given,
`as mandated by the as-needed treatment strategy. In this
`study, we are not able to directly compare monthly dosing
`with a less frequent variable regimen; however, as an ap-
`proximation of such a comparison, we distinguished those
`eyes given injections at an average spacing of 2 months or
`less over the follow-up period from the rest of the study
`population. Comparison of these groups revealed a greater
`mean VA improvement (⬎2 lines difference) and a lower
`proportion with moderate VA loss in the group receiving
`injections at an average interval of 2 months or less (Figs 3
`and 4). In previously reported subgroup analyses of the
`MARINA trial data, higher baseline VA, larger CNV, and
`increasing age were found to be associated with less vis-
`ion gain and greater vision loss in ranibizumab-treated
`groups.13 In our population, we found no such association
`between age and VA change. There was a weak to moderate
`association between lower baseline VA and VA improve-
`ment at month 3, month 6, and last follow-up (Spearman
`r ⫽ 0.27, P ⫽ 0.003; r ⫽ 0.35, P⬍0.0001; r ⫽ 0.26; P ⫽
`0.0026, respectively). However, there was no significant
`difference in age, baseline CNV size, or baseline VA be-
`tween the ⬍2 months and ⬎2 months mean injection inter-
`val subgroups. In fact, the small difference in baseline VA,
`although not statistically significant (P ⫽ 0.15), actually
`favored the ⬎2 months injection interval group, even
`though this group had a less favorable visual outcome in our
`
`1745
`
`Mylan Exhibit 1094
`Mylan v. Regeneron, IPR2021-00880
`Page 6
`
`

`

`Ophthalmology Volume 116, Number 9, September 2009
`
`study. We were also unable to identify any morphologic
`feature on baseline FA or OCT that was different between
`these 2 groups.
`One implication of these findings is that a strictly as-
`needed treatment strategy based on serial OCT scans may
`lead to undertreatment with anti-VEGF therapies. This is
`similar to the findings of the EXCITE trial in which the
`monthly treated patients fared better than the quarterly
`treated groups. An important question that remains to be
`answered is whether continued anti-VEGF injections in
`eyes with apparent resolution of fluid on OCT may lead to
`continued VA improvement. Although some have chosen
`fixed monthly dosing until more definitive data on alterna-
`tive regimens emerge, there is at present no consensus on
`optimal treatment and monitoring strategies, and many cli-
`nicians prefer a variable dosing regimen to avoid unneces-
`sary overtreatment. The conceptual disadvantage of an as-
`needed treatment strategy is its reliance on a deterioration
`event (e.g., recurrence of fluid, clinically apparent hemor-
`rhage, VA loss) as the trigger