A Phase IIIb Study to Evaluate the Safety
`of Ranibizumab in Subjects with
`Neovascular Age-related Macular
`Degeneration
`
`David S. Boyer, MD,1 Jeffrey S. Heier, MD,2 David M. Brown, MD,3 Steven F. Francom, PhD,4
`Tsontcho Ianchulev, MD,4 Roman G. Rubio, MD4
`
`Objective: To evaluate the safety and efficacy of intravitreal ranibizumab in a large population of subjects
`with neovascular age-related macular degeneration (AMD).
`Design: Twelve-month randomized (cohort 1) or open-label (cohort 2) multicenter clinical trial.
`Participants: A total of 4300 subjects with angiographically determined subfoveal choroidal neovascular-
`ization (CNV) secondary to AMD.
`Methods: Cohort 1 subjects were randomized 1:1 to receive 0.3 mg (n ⫽ 1169) or 0.5 mg (n ⫽ 1209)
`intravitreal ranibizumab for 3 monthly loading doses. Dose groups were stratified by AMD treatment history
`(treatment-naïve vs. previously treated). Cohort 1 subjects were retreated on the basis of optical coherence
`tomography (OCT) or visual acuity (VA) criteria. Cohort 2 subjects (n ⫽ 1922) received an initial intravitreal dose
`of 0.5 mg ranibizumab and were retreated at physician discretion. Safety was evaluated at all visits.
`Main Outcome Measures: Safety outcomes included the incidence of ocular and nonocular adverse
`events (AEs) and serious adverse events (SAEs). Efficacy outcomes included changes in best-corrected VA
`over time.
`Results: Some 81.7% of cohort 1 subjects and 49.9% of cohort 2 subjects completed the 12-month study.
`The average total number of ranibizumab injections was 4.9 for cohort 1 and 3.6 for cohort 2. The incidence of
`vascular and nonvascular deaths during the 12-month study was 0.9% and 0.7% in the cohort 1 0.3 mg group,
`0.8% and 1.5% in the cohort 1 0.5 mg group, and 0.7% and 0.9% in cohort 2, respectively. The incidence of
`death due to unknown cause was 0.1% in both cohort 1 dose groups and cohort 2. The number of vascular
`deaths and deaths due to unknown cause did not differ across cohorts or dose groups. Stroke rates were 0.7%,
`1.2%, and 0.6% in the 0.3 mg and 0.5 mg groups and cohort 2, respectively. At month 12, cohort 1 treatment-
`naïve subjects had gained an average of 0.5 (0.3 mg) and 2.3 (0.5 mg) VA letters and previously treated subjects
`had gained 1.7 (0.3 mg) and 2.3 (0.5 mg) VA letters.
`Conclusions:
`Intravitreal ranibizumab was safe and well tolerated in a large population of subjects with
`neovascular AMD. Ranibizumab had a beneficial effect on VA. Future investigations will seek to establish optimal
`dosing regimens for persons with neovascular AMD.
`Financial Disclosure(s): Proprietary or commercial disclosure may be found after the references. Ophthalmology
`2009;116:1731–1739 © 2009 by the American Academy of Ophthalmology.
`
`Neovascular age-related macular degeneration (AMD) is
`characterized by new vessel growth and leakage in the
`choroidal vascular network beneath the macula, with
`extension and leakage into the subretinal space. Although
`the pathologic events that precede choroidal neovascu-
`larization (CNV) are not clearly understood, disrupting
`the activity of vascular endothelial growth factor A
`(VEGF-A), a diffusible cytokine that promotes angiogen-
`esis and vascular permeability, effectively treats CNV
`secondary to AMD.
`Ranibizumab (LUCENTIS, Genentech, Inc., South San
`Francisco, CA) is a recombinant, humanized monoclonal
`antibody antigen-binding fragment (Fab) that neutralizes all
`active forms of VEGF-A. In 2 pivotal phase III trials—
`
`Minimally Classic/Occult Trial of the Anti-VEGF Antibody
`Ranibizumab in the Treatment of Neovascular Age-Related
`Macular Degeneration (MARINA)1 and Anti-Vascular
`Endothelial Growth Factor (VEGF) Antibody for the
`Treatment of Predominantly Classic Choroidal Neovas-
`cularization (CNV) in Age-related Macular Degeneration
`(ANCHOR)2—monthly intravitreal injections of 0.3 mg or
`0.5 mg ranibizumab not only prevented vision loss but also
`improved visual acuity (VA) in patients with minimally
`classic or occult without classic and predominantly classic
`CNV, respectively. In those studies, ranibizumab treatment
`was associated with a low rate of serious adverse events
`(SAEs), including those attributable to systemic VEGF
`inhibition.
`
`© 2009 by the American Academy of Ophthalmology
`Published by Elsevier Inc.
`
`ISSN 0161-6420/09/$–see front matter
`doi:10.1016/j.ophtha.2009.05.024
`
`1731
`
`Regeneron Exhibit 2032
`Page 01 of 09
`
`

`

`Ophthalmology Volume 116, Number 9, September 2009
`
`The Safety Assessment of Intravitreous Lucentis for
`AMD (SAILOR) study was a phase IIIb follow-up study
`to the MARINA and ANCHOR studies to evaluate the
`long-term safety and efficacy of ranibizumab in a large
`population of subjects with all subtypes (minimally clas-
`sic, occult without classic, and predominantly classic) of
`neovascular AMD. SAILOR included more than 5 times
`as many ranibizumab-treated subjects as the MARINA
`and ANCHOR studies combined. Thus, it is the largest
`multicenter randomized study to date to evaluate safety
`and efficacy outcomes of anti-VEGF treatment in wet
`AMD, and it
`is the only phase III study to examine
`individualized, criteria-based retreatment.
`
`Materials and Methods
`
`SAILOR was a 12-month, multicenter, phase IIIb study intended
`to further characterize the safety and efficacy profiles of intravitreal
`ranibizumab. Protocols were approved by the institutional review
`board at each study site, and the study was conducted according to the
`International Conference on Harmonisation E6 Guideline for
`Good Clinical Practice and any national requirements. All sub-
`jects provided informed consent before participation in the
`study. The SAILOR study is registered at www.clinicaltrials.
`gov (NCT00251459; accessed February 5, 2009).
`Two study cohorts were enrolled. Cohort 1 subjects were
`randomized 1:1 to receive 0.3 mg or 0.5 mg intravitreal ranibi-
`zumab. Cohort 2 subjects received open-label 0.5 mg intravitreal
`ranibizumab. Eligible subjects were ⱖ50 years of age with 20/40
`to 20/400 (Snellen equivalent) best-corrected VA in the study eye.
`Cohort 1 VA was assessed with the Early Treatment Diabetic
`Retinopathy Study (ETDRS) chart. In the interest of conserving
`time and resources, VA for cohort 2 (under a less rigorous treat-
`ment and assessment schedule) was assessed using Snellen charts.
`All subjects had angiographically determined subfoveal CNV
`(minimally classic, occult without classic, predominantly classic)
`secondary to AMD (as determined by the investigating physician),
`with evidence of recent disease progression defined by any of the
`following: loss of ⱖ5 ETDRS letters (or ⱖ1 Snellen line) within
`6 months before study initiation (i.e., day 0); 10% increase in the
`CNV lesion area determined by comparing a fluorescein angio-
`gram performed within 1 month before day 0 with an angiogram
`performed within 6 months before day 0; subretinal hemorrhage
`associated with CNV within 1 month before day 0; or classic CNV
`comprising ⬎50% of the CNV lesion area.
`Key exclusion criteria included verteporfin photodynamic ther-
`apy, pegaptanib sodium, or other AMD therapy within 30 days
`before day 0; previous submacular surgery or other surgical inter-
`vention for AMD in the study eye; participation in an investiga-
`tional drug (except vitamins and minerals) study within 30 days
`before day 0; previous participation in a ranibizumab clinical trial;
`intravitreal administration of bevacizumab within 30 days before
`day 0; or current use of systemic anti-VEGF agents. Also excluded
`were subjects with fibrosis or atrophy involving the foveal center
`of the treated eye in the absence of a new lesion; CNV in either eye
`due to other causes, such as ocular histoplasmosis, trauma, or
`pathologic myopia; a tear in the retinal pigment epithelium of the
`study eye involving the macula; or any current intraocular condi-
`tion in the study eye (e.g., cataract or diabetic retinopathy) that, in
`the investigating physician’s opinion, would require medical or
`surgical intervention during the 12-month study period or, if al-
`lowed to progress untreated, would likely contribute to the loss of
`at least 2 Snellen equivalent lines of VA over the 12-month study
`
`1732
`
`Figure 1. Study treatment and assessments. Cohort 1 subjects received 3
`loading doses of ranibizumab and were retreated on the basis of VA (⬎5
`letter decrease in VA from highest score at prior visits) or VA and/or OCT
`(⬎100 ␮m increase in CFT from the lowest measurement at prior visits)
`criteria. Cohort 2 subjects received 1 dose of ranibizumab on day 1 and
`were retreated at physician discretion. CFT ⫽ central foveal thickness;
`OCT ⫽ optical coherence tomography; VA ⫽ visual acuity.
`
`period. Subjects with a history of cardiovascular disease were not
`excluded if their disease was controlled.
`Cohort 1 subjects were randomized 1:1 to receive 0.3 mg or 0.5
`mg intravitreal ranibizumab. To prevent bias in reporting AEs,
`subjects were masked to treatment dose. (Because SAILOR was
`not designed with efficacy as an objective, physicians and study
`monitors were not masked.) Randomization was stratified accord-
`ing to treatment history. “Previously treated” subjects had previ-
`ously received treatment AMD. “Treatment-naïve” subjects were
`newly diagnosed with neovascular AMD. Cohort 1 subjects
`received 3 monthly loading doses of intravitreal ranibizumab (day
`0, month 1, and month 2) with scheduled follow-up visits at
`months 3, 6, 9, and 12 (Fig 1). If, at any time, the investigating
`physician believed that the between-visit interval was too long for
`a patient to go without being assessed, an unscheduled visit could
`occur. After the 3 loading doses, retreatment was based on (1) VA
`(a ⬎5 ETDRS letter decrease in VA compared with the highest
`VA score at any prior scheduled visit) or (2) VA (same as above)
`and/or optical coherence tomography (OCT) (a ⬎100-␮m increase
`in central foveal thickness [CFT] compared with the lowest mea-
`surement at any previous scheduled study visit, with intraretinal or
`subretinal fluid present). Thus, OCT assessment was required only
`for retreatment option 2, in which case OCT data were consistently
`obtained at all study visits. Retreatment was to occur no more
`frequently than every 30 days. Before randomization, the investi-
`gating physician selected the retreatment criterion for each subject
`that was to be used throughout the study.
`Cohort 1 subjects were evaluated with a full ocular exami-
`nation and best-corrected VA (ETDRS chart at a distance of
`4 m) and safety assessments on day 0 and at all scheduled
`(months 1, 2, 3, 6, 9, and 12) visits. Visual acuity assessments
`were required at unscheduled visits if a subject was being
`evaluated for retreatment. Safety assessments were required at
`all unscheduled visits.
`Cohort 2 included both previously treated and treatment-
`naïve subjects. Subjects received 0.5 mg of ranibizumab, with
`an initial injection on day 0 and retreatment at the investigating
`physician’s discretion, no more frequently than every 30 days.
`Cohort 2 subjects were evaluated for Snellen VA at day 0 and
`months 6 and 12. At unscheduled visits, VA was assessed at the
`investigating physician’s discretion. Serious adverse events and
`adverse events (AEs) were assessed at scheduled and unsched-
`uled visits, with formal safety assessments scheduled for
`months 6 and 12.
`
`Regeneron Exhibit 2032
`Page 02 of 09
`
`

`

`Boyer et al
`
`䡠 Ranibizumab Safety in Wet AMD
`
`Adverse events included any unfavorable or unintended sign,
`symptom, or disease temporally associated with use of study drug
`or other protocol-imposed intervention. An AE was classified as an
`SAE if it caused or led to death, required or prolonged subject
`hospitalization, resulted in persistent or significant disability or
`incapacitation, or was considered to be a significant medical event
`by the investigating physician.
`One eye per subject (i.e., the study eye) was treated. After
`thoroughly cleansing the lid, lashes, periorbital area, and conjunc-
`tiva with povidone iodine, local anesthesia and antimicrobials
`(ofloxacin ophthalmic solution, trimethoprim-polymyxin B oph-
`thalmic solution, moxifloxacin ophthalmic solution, or gatifloxacin
`ophthalmic solution) were administered to the study eye. A 30-
`gauge, 0.5-inch needle attached to a low-volume syringe contain-
`ing 50 ␮L of ranibizumab solution was inserted through the
`conjunctiva and sclera, 3.5 to 4.0 mm posterior to the limbus,
`avoiding the horizontal meridian and aiming toward the center of
`the globe. The injection volume was delivered slowly. The needle
`was slowly removed, ensuring that all drug solution was in the eye.
`Immediately after the injection, antimicrobial drops were admin-
`istered, and the subject was instructed to self-administer antimi-
`crobial drops 4 times daily for 3 days. The study eye was assessed
`with a finger count test and intraocular pressure within 15 and 70
`minutes, respectively, of the ranibizumab injection.
`The primary safety end point for cohort 1 was incidence of
`ocular and nonocular SAEs evaluated through month 12. A sec-
`ondary safety end point was incidence of ocular and nonocular
`AEs evaluated through month 12. Efficacy end points for cohort 1
`included change from baseline VA, proportion of subjects who
`gained ⱖ15 VA letters from baseline, and change from baseline
`CFT across the study period.
`The primary safety end points for cohort 2 were the inci-
`dence of ocular and nonocular SAEs and AEs evaluated through
`month 12. Efficacy outcomes for cohort 2 included median
`change in Snellen VA from baseline and the proportion of
`subjects with Snellen 20/200 or worse at baseline compared
`with months 6 and 12.
`
`Statistical Analysis
`
`Safety and efficacy analyses included all subjects who received at
`least 1 injection of ranibizumab. Incidence of ocular and nonocular
`SAEs and AEs and 95% 2-sided confidence intervals for key SAEs
`were determined for both cohorts and each dose group. No formal
`hypothesis testing was conducted to compare cohorts, dose groups,
`or treatment-naïve and previously treated subjects. A sample of
`2378 cohort 1 subjects and 1922 cohort 2 subjects was considered
`sufficient to estimate rates of uncommon SAEs and AEs.
`Efficacy results for cohort 1 were stratified by dose group and
`treatment history. Estimated proportions were obtained for dichot-
`omous end points. Continuous end points were evaluated using
`descriptive statistics, including mean, median, standard deviation,
`standard error, and range.
`To further evaluate stroke rates across cohorts and dose groups,
`each subject’s medical history was reviewed, and subjects were
`classified by preexisting conditions that may have been associated
`with the incidence of stroke during the 12-month study. These
`included prior stroke, myocardial infarction (MI), hypertension,
`transient ischemic attack, coronary artery disease, arrhythmias,
`valve malfunction, congestive heart failure, angioplasty, deep vein
`thrombosis, diabetes, endocardectomy, cardiac inflammation, prior
`stent, and use of aspirin, lipid-lowering drugs, anticoagulants, or
`platelet aggregation inhibitors. A univariate Cox proportional haz-
`ard regression model was used to identify which of those were
`significant (i.e., Pⱕ0.05) risk factors for stroke in SAILOR. In
`
`addition, models that included the interaction of dose with each of
`the significant risk factors were fit separately.
`
`Missing Data
`Missing data were not imputed for safety end points. For cohort 1,
`missing values for efficacy end points were imputed using the
`last-observation-carried-forward method. For cohort 2, missing
`Snellen values were not imputed.
`
`Results
`
`From November 2005 to June 30, 2006 (when ranibizumab was
`approved for the treatment of neovascular AMD by the Food and
`Drug Administration), 2378 cohort 1 subjects were randomly
`assigned to receive 0.3 mg (n ⫽ 1169) or 0.5 mg (n ⫽ 1209)
`intravitreal ranibizumab at 105 US centers. Cohort 1 subjects had
`an average age of 79 years, and 59% were female (Table 1).
`Approximately 60% of cohort 1 subjects in each dose group had
`been previously treated for AMD. The types of previous treatment
`were similar across dose groups and included photodynamic ther-
`apy (33%), intravitreal pegaptanib sodium (30%), intravitreal tri-
`amcinolone acetonide (17%), and laser photocoagulation (10%).
`Investigating physicians elected to use the VA plus OCT retreat-
`ment criterion for approximately 81% of the subjects in each dose
`group.
`Previously treated and treatment-naïve subjects had similar
`baseline ocular characteristics, with the exception that previously
`treated subjects had a longer time since first diagnosis and lower
`baseline VA (Table 2). Approximately 18% of cohort 1 subjects in
`each dose group discontinued the study before the month 12 visit
`(Table 3). Baseline ocular characteristics of subjects who com-
`
`Table 1. Subject Baseline Characteristics
`
`Characteristic
`
`Age (yrs)
`Mean ⫾ SD
`Range
`Sex
`Female
`Race
`Caucasian
`AMD treatment
`history
`Treatment naïve
`Previously treated
`Retreatment criteria
`VA
`VA plus OCT
`Systolic BP
`Mean ⫾ SD
`Range
`Diastolic BP
`Mean ⫾ SD
`Range
`
`Cohort 1
`
`0.3 mg
`(n ⫽ 1169)
`
`0.5 mg
`(n ⫽ 1209)
`
`Cohort 2
`0.5 mg
`(n ⫽ 1922)
`
`78.7⫾7.6
`51–97
`
`78.7⫾8.6
`52–101
`
`78.7⫾8.1
`45–99
`
`59.9
`
`96.6
`
`39.5
`60.5
`
`19.3
`80.7
`
`58.1
`
`97.1
`
`40.5
`59.5
`
`18.4
`81.6
`
`137.4⫾17.3
`90–213
`
`137.8⫾18.0
`80–220
`
`76.2⫾9.7
`48–118
`
`77.0⫾9.7
`48–110
`
`61.6
`
`96.2
`
`—
`—
`
`—
`—
`
`—
`—
`
`—
`—
`
`AMD ⫽ age-related macular degeneration; BP ⫽ blood pressure; OCT ⫽
`optical coherence tomography; SD ⫽ standard deviation; VA ⫽ visual
`acuity.
`Values are percentages except where otherwise noted.
`
`1733
`
`Regeneron Exhibit 2032
`Page 03 of 09
`
`

`

`Ophthalmology Volume 116, Number 9, September 2009
`
`Table 2. Baseline Ocular Characteristics
`
`Treatment Naive
`
`Cohort 1
`
`0.3 mg
`(n ⫽ 462)
`79.9⫾7.9
`0.3⫾1.4
`
`32.0
`19.7
`45.5
`
`55.0⫾12.5
`
`20/80
`12.2
`312⫾104
`15.3⫾3.2
`
`0.5 mg
`(n ⫽ 490)
`75.8⫾8.0
`0.3⫾0.7
`
`29.4
`20.2
`48.6
`
`48.9⫾13.8
`
`20/80
`15.0
`322⫾116
`15.3⫾3.2
`
`Previously Treated
`0.3 mg
`0.5 mg
`(n ⫽ 707)
`(n ⫽ 719)
`79.9⫾7.5
`79.9⫾7.5
`1.4⫾2.0
`1.3⫾1.7
`
`30.6
`26.2
`38.6
`
`53.8⫾13.8
`
`20/100
`22.9
`315⫾113
`15.7⫾3.3
`
`31.7
`23.5
`40.6
`
`50.0⫾14.3
`
`20/100
`23.0
`310⫾113
`15.4⫾3.4
`
`Cohort 2
`0.5 mg
`(n ⫽ 1922)
`
`—
`—
`
`—
`—
`—
`
`—
`
`20/100
`39
`—
`—
`
`Age at diagnosis (yrs)
`Time since diagnosis (yrs)
`CNV type (%)
`Predominantly classic
`Minimally classic
`Occult without classic
`VA
`ETDRS letters
`Snellen
`Median
`20/200 or worse (%)
`Central foveal thickness (␮m)
`Intraocular pressure (mmHg)
`
`CNV ⫽ choroidal neovascularization; ETDRS ⫽ Early Treatment Diabetic Retinopathy Study; VA ⫽ visual acuity.
`Values are mean ⫾ standard deviation except where otherwise noted.
`
`pleted the study and those who discontinued were similar. All
`cohort 1 subjects received their assigned dose of ranibizumab on
`day 0, and approximately 96% of cohort 1 subjects received their
`assigned dose at months 1 and 2 (Fig 2). Cohort 1 subjects received
`an average of 4.6 injections during the 12-month study (the pro-
`tocol required 3 initial injections). The average number of visits
`was 8.8 (the protocol required 7 scheduled visits). During months
`that visits were not scheduled (i.e., months 4, 5, 7, 8, 10, and 11),
`approximately 40% of the subjects made unscheduled visits, and
`approximately 16% of those subjects received an injection of
`ranibizumab at the unscheduled visit (relative to the number of
`subjects remaining in the study that month) (Fig 2).
`From March 2006 to June 30, 2006, 1922 cohort 2 subjects
`were enrolled at 104 US centers and received 0.5 mg intravitreal
`ranibizumab (Table 1). Approximately 50% of cohort 2 subjects
`discontinued the study before the month 12 visit (Table 3). All
`cohort 2 subjects received the protocol-required injection on day 0
`and received an average of 3.6 injections during the 12-month
`study (the protocol required 1 injection). The average number of
`visits for cohort 2 subjects was 4.9 (the protocol required 3
`scheduled visits). During months that visits were not required
`
`Table 3. Reasons for Discontinuation
`
`Cohort 1
`0.3 mg
`0.5 mg
`(n ⫽ 1169)
`(n ⫽ 1209)
`
`Cohort 2
`0.5 mg
`(n ⫽ 1922)
`
`18.6
`
`18.0
`
`1.7
`2.6
`0.7
`6.7
`3.4
`0.2
`0.6
`2.7
`
`0.1
`
`2.3
`2.2
`0.9
`5.8
`2.8
`0.1
`0.9
`3.1
`
`0
`
`50.1
`
`1.5
`1.8
`2.0
`29.0
`9.4
`0.3
`0.9
`5.3
`
`0
`
`Discontinued early (%)
`Reason for early
`discontinuation (%)
`Death
`Adverse event
`Loss to follow-up
`Subject decision
`Physician decision
`Sponsor decision
`Subject noncompliance
`Subject’s condition mandated
`other therapeutic
`intervention
`Reason not provided
`
`1734
`
`(i.e., all but months 6 and 12), the percentage of subjects who
`remained in the study that made unscheduled visits ranged from
`65% at month 2 to 17.4% at month 11. The percentage of
`subjects receiving injections ranged from 64% at month 2 to
`16.5% at month 11.
`
`Figure 2. Visits and treatment. The percentage of cohort 1 (upper) and
`cohort 2 (lower) patients making visits and receiving ranibizumab treat-
`ment during each month of the 12-month study are shown. Cohort 1 visits
`were scheduled for day 0 and months 1, 2, 3, 6, 9, and 12. Cohort 2 visits
`were scheduled for day 0 and months 6 and 12. Data from cohort 1 0.3 and
`0.5 mg dose groups are combined. Values are based on the percentage of
`subjects remaining in the study at each time point. Treatment received at
`month 12 was in violation of the protocol.
`
`Regeneron Exhibit 2032
`Page 04 of 09
`
`

`

`Boyer et al
`
`䡠 Ranibizumab Safety in Wet AMD
`
`Table 4. Key Ocular Serious Adverse Events
`
`Event, %
`
`Presumed endophthalmitis*
`Uveitis
`Retinal detachment
`Retinal tear
`Retinal hemorrhage
`Detachment of retinal
`pigment epithelium
`Vitreous hemorrhage
`Cataract
`
`Cohort 1
`0.3 mg
`0.5 mg
`(n ⫽ 1169)
`(n ⫽ 1209)
`
`Cohort 2
`0.5 mg
`(n ⫽ 1922)
`
`0.2
`0.1
`0.1
`0
`0.9
`0
`
`0.3
`0.1
`
`0.4
`0.2
`0
`0.1
`0.9
`0.2
`
`0.1
`0.1
`
`0.1
`0
`0.1
`0
`0.3
`0.1
`
`0.2
`0.1
`
`*Includes 2 cases of uveitis and 1 case of iridocyclitis that were treated
`with antibiotics.
`
`Safety
`
`Ocular safety. The rates of individual key ocular SAEs in cohort
`1 were ⬍1% and similar across dose groups (Table 4). Two
`subjects (0.2%) in the 0.3 mg group and 5 subjects (0.4%) in the
`0.5 mg group developed endophthalmitis or presumed endoph-
`thalmitis (i.e., ocular infection treated with antibiotics). One sub-
`ject in each cohort 1 dose group had a serious cataract event. The
`rates of individual key ocular SAEs in cohort 2 were ⬍1%. One
`cohort 2 subject developed endophthalmitis, and 1 subject had a
`serious cataract event (Table 4).
`The incidence of ocular inflammation AEs, including iritis,
`uveitis, vitritis, and iridocyclitis, was 1.0% in the 0.3 mg group,
`1.5% in the 0.5 mg group, and 0.5% in cohort 2. The overall
`incidence of cataract AEs was 5.4% in the 0.3 mg group, 6.0%
`in the 0.5 mg group, and 2.8% in cohort 2, and was similar when
`broken down by nuclear, subcapsular, and cortical subtypes.
`Nonocular safety. The rates of key nonocular SAEs were sim-
`ilar across cohort 1 dose groups (Fig 3; Table 5). Nonvascular
`death, stroke, and hemorrhage rates were numerically higher in the
`0.5 mg group. Eight subjects (0.7%) in the 0.3 mg group and 15
`subjects (1.2%) in the 0.5 mg group had a stroke during the
`
`Figure 3. Key nonocular SAEs. The rates of individual events are de-
`picted as point estimates with 2-sided Blyth-Still-Casella 95% confidence
`intervals. Antiplatelet Trialists’ Collaboration ATEs include vascular
`deaths and deaths due to unknown cause, nonfatal MI, and nonfatal
`stroke. APTC ⫽ Antiplatelet Trialists’ Collaboration; ATE ⫽ arterial
`thromboembolic event; SAE ⫽ serious adverse events.
`
`Table 5. Nonocular Adverse Events Potentially Related to
`Anti-Vascular Endothelial Growth Factor Therapy
`
`Classification, %
`
`Arterial thromboembolic
`events
`All
`Serious
`Hypertension
`All
`Serious
`Nonocular hemorrhage
`All
`Serious
`Proteinuria
`All
`Serious
`Other
`All
`Serious
`
`Cohort 1
`0.3 mg
`0.5 mg
`(n ⫽ 1169)
`(n ⫽ 1209)
`
`Cohort 2
`0.5 mg
`(n ⫽ 1922)
`
`3.8
`2.5
`
`9.0
`0.1
`
`2.9
`0.9
`
`0.1
`0
`
`0.7
`0.3
`
`4.1
`3.1
`
`10.3
`0.1
`
`3.1
`1.5
`
`0
`0
`
`0.4
`0.2
`
`2.4
`1.6
`
`3.0
`0
`
`1.4
`0.6
`
`0
`0
`
`0.1
`0.1
`
`VEGF ⫽ vascular endothelial growth factor.
`
`12-month study period. The incidence of MI and Antiplatelet
`Trialists’ Collaboration (APTC)3 arterial thromboembolic events
`(ATEs), which include vascular death and death of unknown
`cause, nonfatal MI, and nonfatal cardiovascular accidents, were
`similar across cohort 1 dose groups.
`Rates of key nonocular SAEs in cohort 2 were generally
`lower than those in cohort 1, which may be a result of under-
`reporting because of the large number of cohort 2 subjects who
`discontinued. The incidence of nonocular AEs potentially re-
`lated to anti-VEGF therapy was low and comparable across
`cohorts and dose groups.
`Prior stroke, history of arrhythmias, and history of congestive
`heart failure were significant risk factors for stroke (Fig 4). Al-
`though the numbers were small, there was a nonstatistically sig-
`nificant trend toward higher incidence of stroke in the cohort 1 0.5
`mg group subjects with a history of stroke. Seven of the 73
`subjects (9.6%) with a history of stroke in the 0.5 mg group
`experienced a stroke during the study compared with 2 of the 73
`subjects (2.7%) with a history of stroke in the 0.3 mg group. None
`of the cohort 2 subjects with a history of stroke experienced a
`stroke during the study (Fig 4).
`Twenty subjects (1.7%) in the cohort 1 0.3 mg group, 29
`subjects (2.4%) in the cohort 1 0.5 mg group, and 33 subjects
`(1.7%) in cohort 2 died during the 12-month study (Table 6). The
`number of vascular deaths and deaths due to unknown cause did
`not differ across cohorts or dose groups.
`
`Efficacy
`Cohort 1 efficacy results were stratified by dose and previous
`treatment for AMD. For all groups, study eye VA increased with
`3 loading doses of ranibizumab (day 0, month 1, month 3) (Fig 5).
`At month 3, treatment-naïve subjects in the 0.3 mg group had
`gained an average of 5.8 VA letters and those in the 0.5 mg group
`had gained an average of 7.0 VA letters. From months 3 to 12, with
`protocol-defined retreatment, VA tended to decrease. At month 12,
`treatment-naïve subjects in the 0.3 mg group had gained an aver-
`age of 0.5 VA letters and those in the 0.5 mg group had gained an
`average of 2.3 letters. A similar pattern was observed for previ-
`
`1735
`
`Regeneron Exhibit 2032
`Page 05 of 09
`
`

`

`Ophthalmology Volume 116, Number 9, September 2009
`
`ously treated subjects. At month 3, previously treated subjects in
`the 0.3 mg group had gained an average of 4.6 VA letters and those
`in the 0.5 mg group had gained an average of 5.8 VA letters. At
`month 12, previously treated subjects in the 0.3 mg group had
`gained an average of 1.7 VA letters and those in the 0.5 mg group
`had gained an average of 2.3 letters.
`In all cohort 1 groups, the proportion of subjects who gained
`ⱖ15 letters from baseline VA increased with 3 loading doses of
`ranibizumab (Fig 6). At month 3, 19.4% of treatment-naïve sub-
`jects in the 0.3 mg group and 20.1% in 0.5 mg group had gained
`ⱖ15 letters. The proportion of those who gained ⱖ15 letters
`tended to be maintained for the duration of the 12-month study,
`with 14.6% of 0.3 mg group subjects and 19.3% of 0.5 mg subjects
`gaining ⱖ15 VA letters at month 12. A similar pattern was
`observed for previously treated subjects. At month 3, 16.0% of
`previously treated subjects in the 0.3 mg group and 18.6% in the
`0.5 mg group had gained ⱖ15 letters; and at month 12, 15.8% of
`0.3 mg group subjects and 16.5% of 0.5 mg group subjects had
`gained ⱖ15 VA letters.
`Study eye CFT of cohort 1 subjects for whom OCT data
`were available decreased with 3 loading doses of ranibizumab,
`increased from months 3 to 6, and remained stable from months
`6 to 12 (Fig 7). For treatment-naïve subjects, CFT had de-
`creased an average of 107.0 ␮m in the 0.3 mg group and 122.0
`␮m in the 0.5 mg group at month 3. At month 12, the average
`decrease from baseline CFT was 72.0 ␮m in the 0.3 mg group
`and 92.0 ␮m in the 0.5 mg group. For previously treated
`subjects, CFT had decreased an average of 98.0 ␮m in the 0.3
`mg group and 108.0 ␮m in the 0.5 mg group at month 3. At
`month 12, the average decrease from baseline CFT was 71.0
`␮m in the 0.3 mg group and 76.0 ␮m in the 0.5 mg group.
`Because of the large number of cohort 2 subjects who discon-
`tinued, the last-observation-carried-forward method was not used
`to impute missing efficacy values, and observed results are re-
`ported. This should be kept in mind when interpreting the results.
`
`Table 6. Cause of Death
`
`Cohort 1
`0.3 mg
`0.5 mg
`(n ⫽ 1169)
`(n ⫽ 1209)
`
`Cohort 2
`0.5 mg
`(n ⫽ 1922)
`
`All deaths, %
`Deaths due to unknown
`cause, %
`Vascular deaths, %
`Cardiovasculara
`Strokeb
`Nonvascular deaths, %
`Respiratory: pneumonia,
`respiratory failure
`pulmonary failure
`pulmonary edema
`Accident, injury, intracranial
`bleed secondary to fall
`Renal failure
`Cancer
`Infection (septic shock,
`sepsis, urosepsis), liver
`failure due to hepatitis
`Postoperative bowel
`obstruction
`Vasculitis
`
`1.7
`0.1
`
`0.9
`0.8
`0.2
`0.7
`0.3
`
`0
`
`0
`0
`0.3
`
`0.1
`
`0
`
`2.4
`0.1
`
`0.8
`0.5
`0.3
`1.5
`0.6
`
`0.2
`
`0.1
`0.4
`0.2
`
`0
`
`0
`
`1.7
`0.1
`
`0.7
`0.7
`0.1
`0.9
`0.5
`
`0.1
`
`0
`0.3
`0
`
`0
`
`0.1
`
`aIncludes ischemic cardiomyopathy, coronary heart disease, cardiac arrest,
`MI, saddle pulmonary embolism, and heart failure.
`bIncludes stroke, acute ischemic stroke, intracerebral hemorrhage, cere-
`brovascular disease, and brain hemorrhage secondary to fall. Three 0.5 mg
`subjects with preexisting cancer had previously received cancer treatment.
`
`Snellen VA in cohort 2 subjects improved from a median of
`20/100 at baseline to 20/80 at months 6 and 12. The proportion of
`subjects with a Snellen equivalent of 20/200 or worse decreased
`from approximately 39% at baseline to 31% at month 6 and 32%
`at month 12.
`
`Discussion
`
`SAILOR is the largest study to date to evaluate safety
`(primary objective) and efficacy (secondary objective) of
`
`Figure 4. Stroke rate by risk factor. Point estimates and 2-sided Blyth-
`Still-Casella 95% confidence intervals for stroke rate when the risk factor
`was present or absent are shown. We evaluated the impact of 21 factors on
`the incidence of stroke. The 5 risk factors that had the greatest effect on
`stroke rates are presented.
`
`Figure 5. Change from baseline VA (cohort 1). For all groups, VA
`increased with 3 loading doses of ranibizumab (day 0, month 1, month 3).
`From months 3 to 12, with protocol-defined retreatment, VA tended to
`decrease. Error bars are ⫾1 standard error. ETDRS ⫽ Early Treatment
`Diabetic Retinopathy Study; VA ⫽ visual acuity.
`
`1736
`
`Regeneron Exhibit 2032
`Page 06 of 09
`
`

`

`Boyer et al
`
`䡠 Ranibizumab Safety in Wet AMD
`
`In SAILOR there was not a difference between doses in
`APTC ATEs overall, which is consistent with our current
`understanding of ranibizumab pharmacology. As a Fab,
`ranibizumab has low systemic bioavailability (⬃1/90,0000
`of intravitreal concentration) and a half-life of only several
`hours (Kubler P, Xu L, Jumbe N, et al. Population pharma-
`cokinetics of ranibizumab in patients with age-related mac-
`ular degeneration. Presented at: American Society of Retina
`Specialists Annual Meeting, December 1–5, 2007; Indian
`Wells, California).
`Certain subgroups of subjects (e.g., those with prior
`cardiovascular accidents) may experience higher rates of
`systemic SAEs. We observed that the incidence of stroke
`was greater for cohort 1 subjects who had a history of
`stroke, congestive heart failure, or arrhythmias. However,
`the low incidence of stroke in SAILOR made it difficult to
`draw meaningful conclusions about the relationship be-
`tween risk factors and stroke. Although the results of clin-
`ical trials cannot be directly compared with epidemiology
`studies in AMD, epidemiology stroke rates can provide a
`reference that aids in understanding stroke rates in SAILOR.
`The annual stroke rate for new-onset neovascular AMD in a
`large sample of Medicare subjects was 3.8%, and the annual
`ischemic stroke rate was 56.4% for those subjects who had
`experienced an ischemic stroke in the year before study
`entry.7 Both of these rates are higher than those observed in
`SAILOR.
`Ranibizumab treatment was associated with a net gain
`in VA in the cohort 1 0.3 mg and 0.5 mg dose groups.
`However, consistent with the results of MARINA and
`ANCHOR, 0.5 mg doses of ranibizumab tended to have a
`slightly greater VA benefit than 0.3 mg doses in subjects
`with neovascular AMD. Ranibizumab also tended to be
`
`Figure 7. Change from baseline CFT (cohort 1). In cohort 1 subjects with
`OCT data, CFT decreased with 3 loading doses of ranibizumab. Central
`foveal thickness then increased from months 3 to 6 and remained stable
`from months 6 to 12. Error bars are ⫾1 standard error. CFT ⫽ central
`foveal thickness; ETDRS ⫽ Early Treatment Diabetic Retinopathy Study;
`OCT ⫽ optical coherence tomography.
`
`1737
`
`Figure 6. Subjects gaining ⱖ15 letters from baseline VA (cohort 1). The
`proportion of cohort 1 subjects who gained ⱖ15 letters increased with 3
`loading doses of ranibizumab and was th