A Phase I Trial of an IV-Administered
`Vascular Endothelial Growth Factor Trap
`for Treatment in Patients with Choroidal
`Neovascularization due to Age-Related
`Macular Degeneration
`
`Quan Dong Nguyen, MD, MSc,1 Syed Mahmood Shah, MD,1 Gulnar Hafiz, MD,1 Edward Quinlan, MD,1
`Jennifer Sung, MD,1 Karen Chu, MS,2 Jesse M. Cedarbaum, MD,2 Peter A. Campochiaro, MD,1
`CLEAR-AMD 1 Study Group*
`
`Objectives: To assess the safety, pharmacokinetics, and biological activity of IV administration of vascular
`endothelial growth factor trap (VEGF Trap), a recombinant protein containing the binding domains of VEGF
`receptors 1 and 2, in patients with neovascular age-related macular degeneration (AMD).
`Design: Randomized, multicenter, placebo-controlled clinical trial.
`Participants: Twenty-five patients were enrolled (11 male, 14 female); 19 received VEGF Trap (0.3 [n ⫽ 7],
`1.0 [n ⫽ 7], or 3.0 mg/kg [n ⫽ 5]), and 6 received a placebo.
`Methods: Patients were randomized to receive a placebo or 0.3-, 1.0-, or 3.0-mg/kg VEGF Trap—a single
`IV dose followed by a 4-week observation period and then 3 doses 2 weeks apart.
`Main Outcome Measures: Safety and biological activity, including change in excess retinal thickness and
`volume assessed by optical coherence tomography and visual acuity (VA) measured by the Early Treatment
`Diabetic Retinopathy Study protocol.
`Results: The majority of adverse events attributable to VEGF Trap were mild to moderate in severity, but 2
`of 5 patients treated with 3.0 mg/kg experienced dose-limiting toxicity (1 with grade 4 hypertension and 1 with
`grade 2 proteinuria); therefore, all patients in the 3.0 mg/kg– dose group were withdrawn from the study. The
`mean percent changes in excess retinal thickness were ⫺12%, ⫺10%, ⫺66%, and ⫺60%, respectively, for the
`placebo and 0.3-, 1.0-, and 3.0-mg/kg groups at day 15 (P⬍0.02 by analysis of covariance [ANCOVA]) and
`⫺5.6%, ⫹47.1%, and ⫺63.3% for the placebo and 0.3- and 1.0-mg/kg groups at day 71 (P⬍0.02, ANCOVA). A
`significant change in VA was not noted in this small study.
`Conclusions: The maximum tolerated dose of IV VEGF Trap in this study population was 1.0 mg/kg. This
`dose resulted in elimination of about 60% of excess retinal thickness after either single or multiple administra-
`tions. Alternative routes of delivery to increase the therapeutic window are being explored. Ophthalmology 2006;
`113:1522–1532 © 2006 by the American Academy of Ophthalmology.
`
`Age-related macular degeneration (AMD) is the most com-
`mon cause of severe vision loss in patients over the age of
`60 years in developed countries.1 It is recognized by the
`presence of drusen, yellow deposits beneath the retina, and
`
`pigmentary changes due to atrophy and proliferation of
`retinal pigment epithelium (RPE) cells. These changes are
`accompanied by gradual death of photoreceptors and mod-
`erate decreases in visual acuity (VA); this constellation of
`
`Originally received: November 4, 2005.
`Manuscript no. 2005-1071.
`Accepted: May 10, 2006.
`1 Wilmer Eye Institute, Johns Hopkins University School of Medicine,
`Baltimore, Maryland.
`2 Regeneron Pharmaceuticals, Tarrytown, New York.
`Supported by a grant from Regeneron Pharmaceuticals.
`Dr Nguyen is a recipient of a K23 Career Development Award (grant
`no.: EY 13552) from the National Eye Institute, Bethesda, Maryland. Dr
`Campochiaro is the George S. and Dolores Dore Eccles Professor of
`
`Ophthalmology and Neuroscience. Dr Campochiaro has acted as a
`consultant for Regeneron, which has been monitored by the conflict of
`interest committee of Johns Hopkins University School of Medicine.
`Ms Chu and Dr Cedarbaum are employees of Regeneron, which has a
`commercial interest in the VEGF Trap.
`
`Correspondence to Peter A. Campochiaro, MD, Maumenee 719, Wilmer
`Eye Institute, Johns Hopkins University School of Medicine, 600 North
`Wolfe Street, Baltimore, MD 21287-9277. E-mail: pcampo@jhmi.edu.
`
`*For Study Group membership, see “Appendix.”
`
`1522.e1
`
`© 2006 by the American Academy of Ophthalmology
`Published by Elsevier Inc.
`
`ISSN 0161-6420/06/$–see front matter
`doi:10.1016/j.ophtha.2006.05.055
`
`
`
`
`Apotex Exhibit 1023
`Page 1 of 14
`
`

`

`Nguyen et al
`
`䡠 VEGF Trap and Choroidal Neovascularization
`
`signs is referred to as atrophic or nonneovascular AMD.
`Patients with nonneovascular AMD are at risk for develop-
`ment of choroidal neovascularization and thereby convert-
`ing to neovascular AMD. Patients with neovascular AMD
`account for only about 10% of patients with AMD, but
`account for the majority of those with severe vision loss.2
`The pathogenic events underlying conversion from non-
`neovascular to neovascular AMD are uncertain, but studies
`in animal models suggest that increased expression of vas-
`cular endothelial growth factor (VEGF) is likely to play a
`critical role. Inhibition of VEGF receptor signaling by sys-
`temic administration of kinase inhibitors3 or a blockade of
`VEGF by intraocular injection of an anti-VEGF antibody
`fragment4 significantly suppresses choroidal neovascular-
`ization in animal models. These data suggest that VEGF is
`an important therapeutic target for treatment of choroidal
`neovascularization.
`Several VEGF antagonists have been developed and tested
`in patients with neovascular AMD. Pegaptanib (Macugen,
`[OSI] Eyetech Pharmaceuticals, Inc., New York, New
`York) is an RNA molecule that binds VEGF165 but not other
`isoforms of VEGF-A. Intraocular injection of pegaptanib
`every 6 weeks for 1 year reduced the percentage of patients
`with classic choroidal neovascularization due to AMD who
`experienced severe loss of vision (loss of ⱖ15 letters) from
`45% in the sham injection group to 30%.5 Six percent of
`patients treated with pegaptanib, compared with 2% in the
`sham injection group, had a substantial improvement in
`vision (gain of ⱖ15 letters). Relative to sham treatment, the
`increase in size of choroidal neovascularization lesions was
`slowed but not stopped. These benefits are modest, but they
`confirmed that VEGF is a therapeutic target in neovascular
`AMD.
`Ranibizumab (Lucentis, Genentech, Inc., South San
`Francisco, CA) is another VEGF antagonist; it is a Fab
`fragment of an antibody that binds all isoforms of VEGF-A.
`Monthly intraocular injections of ranibizumab in AMD
`patients with occult or minimally classic subfoveal choroi-
`dal neovascularization reduced the percentage of patients
`with severe loss of vision over the course of a year from
`38% in the sham injection group to 5%, and the percentage
`of patients who experienced substantial improvement in
`vision increased from 4.6% to 34% (reported at the meeting
`of the American Society of Retina Specialists, July 2005,
`Montreal, Canada). These data suggest that antagonism of
`all isoforms of VEGF-A in AMD patients with choroidal
`neovascularization can result in stabilization of vision in the
`majority of patients and substantial improvement in vision
`in about a third of patients. Preliminary reports using ranibi-
`zumab in patients with classic subfoveal choroidal neovas-
`cularization have been equally impressive and confirm that
`VEGF-A is a very important target in the treatment of
`neovascular AMD, but suggest several questions. Are the
`superior results with intravitreous injection of ranibizumab
`compared with those with pegaptanib due to the inhibition
`of all VEGF-A isoforms, compared with inhibition of only
`VEGF165; superior pharmacokinetics; a combination of
`both; or some other reason?
`If the superiority of ranibizumab over pegatanib is due to
`its ability to neutralize all of the VEGF-A isoforms as
`
`opposed to just VEGF165, then perhaps neutralizing other
`VEGF family members in addition to VEGF-A would pro-
`vide an even greater benefit. Placental growth factor is a
`VEGF family member that contributes to ocular neovascu-
`larization and excessive vascular permeability, providing a
`rationale for targeting multiple family members and not just
`VEGF-A.6 Vascular endothelial growth factor receptors
`bind multiple members of the VEGF family; VEGF receptor
`1 binds VEGF-A and placental growth factor, and VEGF
`receptor 2 binds VEGF-A, VEGF-B, and VEGF-C. A sol-
`uble form of VEGF receptor 1 consisting of just the extra-
`cellular domain—the receptor without the transmembrane
`or intracellular portions of the molecule—is normally pro-
`duced in the body and seems to be a component of the
`endogenous system that modulates new vessel growth.
`Gene therapy resulting in elevated levels of soluble VEGF
`receptor 1 inhibits ocular neovascularization in animal mod-
`els and has antiangiogenic activity.7–15
`The VEGF Trap is a recombinant soluble VEGF receptor
`protein in which the binding domains of VEGF receptors 1
`and 2 are combined with the Fc portion of immunoglobulin
`G. The receptor portion of the molecule has a very high
`⬍ 1 pmol/l), placental
`affinity for all VEGF-A isoforms (Kd
`growth factor 1 and 2, and VEGF-B, VEGF-C, and VEGF-D.
`So VEGF Trap is distinguished from ranibizumab by its
`higher potency for neutralization of all VEGF-A isoforms
`and its ability to inhibit other related proangiogenic and
`pro-permeability VEGF family members. The Fc portion
`slows clearance by conferring the long circulating half-life
`of an antibody to the molecule.16 Either systemic or intra-
`vitreous administration of VEGF Trap strongly suppressed
`laser-induced choroidal neovascularization in mice17 and in
`primates (Association for Research in Vision and Ophthal-
`mology abstract no. 3258, 2005). The broader activity and
`the higher affinity of VEGF Trap are theoretical advantages
`over ranibizumab, but whether theoretical advantages trans-
`late into practical advantages can be determined only by
`clinical trials.
`Although intravitreous injection of a VEGF antagonist
`has the benefit of limiting systemic exposure, it has the
`disadvantage of producing large fluctuations in intraocular
`levels, which cannot be easily measured, making it impos-
`sible to determine effective tissue levels. It is possible to
`measure plasma levels after intravascular administration,
`and because there is no barrier between the systemic circu-
`lation and the choroid, in this setting plasma levels approx-
`imate tissue levels, making pharmacodynamic analyses pos-
`sible. Correlating plasma levels with toxicity is also
`extremely useful because it provides knowledge regarding
`plasma levels that should be avoided. This allows the ratio-
`nal design of alternative modes of delivery by measuring
`plasma levels that occur after local administration and mak-
`ing sure they fall below toxic levels. For this reason, prob-
`ing the safety of an agent after systemic administration
`provides important information regardless of the ultimate
`mode of delivery. Systemic administration of bevacizumab
`(Avastin, Genentech) in combination with antimetabolites
`has been shown to cause hypertension and proteinuria and is
`associated with an increased risk of
`thromboembolic
`events.18 In an oncology trial in which patients receive
`
`1522.e2
`
`
`
`
`Apotex Exhibit 1023
`Page 2 of 14
`
`

`

`Ophthalmology Volume 113, Number 9, September 2006
`
`multiple subcutaneous injections of VEGF Trap, only hy-
`pertension and proteinuria have been identified as definite
`drug-related side effects. Based in part on this reasonable
`safety profile, we performed a dose-ranging, placebo-con-
`trolled, randomized trial investigating the safety, pharma-
`cokinetics, and biological activity of
`IV-administered
`VEGF Trap in patients with neovascular AMD.
`
`Materials and Methods
`
`Ethical Considerations
`The study was conducted in compliance with the Declaration of
`Helsinki, US Code 21 of Federal Regulations, and the Harmonized
`Tripartite Guidelines for Good Clinical Practice (1996). The study
`was reviewed and approved by the Western Institutional Review
`Board for some centers and by local institutional review boards for
`others. There were 3 participating centers: MacPherson Retina
`Center at Baylor College of Medicine, Houston, Texas; New York
`Eye and Ear Infirmary, New York, New York; and Wilmer Eye
`Institute of the Johns Hopkins University School of Medicine,
`Baltimore, Maryland. Each study subject had comprehensive dis-
`cussions with one of the investigators and gave written informed
`consent before study entry.
`
`Study Design
`The study was a randomized, placebo-controlled, dose escalation
`trial in subjects with subfoveal choroidal neovascularization due to
`neovascular AMD. Three dose levels of VEGF Trap were inves-
`tigated (0.3, 1.0, and 3.0 mg/kg). In each cohort, subjects were
`randomized in a 3:1 ratio (drug:placebo). After enrollment in the
`0.3-mg/kg cohort was completed, there was a 2-week waiting
`period before the 1.0-mg/kg cohort was begun. There was also a
`2-week waiting period after the 1.0-mg/kg cohort was enrolled
`before enrollment of the 3.0-mg/kg cohort was begun.
`
`Study Population
`The main inclusion criteria for the study were (1) male or female
`(any ethnicity), 50 years or older; (2) diagnosis of neovascular
`AMD in the study eye with leaking subfoveal choroidal neovascu-
`larization ⱕ 12 optic disc areas (measured according to the protocol
`of the Macular Photocoagulation Study1); (3) best-corrected VA
`(BCVA) score of 20/40 or worse; (4) clear ocular media and adequate
`pupillary dilation (able to dilate pupils to ⱖ4 mm using standard
`mydriatics) to permit good stereoscopic fundus photography; and
`(5) retinal thickness ⱖ 250 ␮m in the macular region as measured by
`optical coherence tomography (OCT).
`Subjects were excluded from the study if they (1) had another
`disease besides neovascular AMD in the study eye that could
`affect vision or safety evaluation, (2) had previous laser photoco-
`agulation to the center of the fovea in the study eye, (3) were
`eligible for photodynamic therapy in the study eye unless they
`refused it, or (4) had intraocular surgery or another treatment in the
`study eye within 3 months of screening. Nor were subjects eligible
`for the study if they had (1) symptomatic or unstable coronary
`artery disease, angina, congestive heart failure, or an arrhythmia
`requiring active medical management within the last 30 days; (2)
`myocardial infarction or treatment for acute congestive heart fail-
`ure within the past 6 months; (3) malignancy other than basal cell
`carcinoma of the skin diagnosed and treated within the last 5 years;
`(4) a history of peripheral vascular disease; (5) blood pressure
`(BP), treated or untreated, ⬎ 140/90 mmHg on at least 3 repeated
`
`1522.e3
`
`determinations on separate days during the 6 weeks before study
`entry; or (6) abnormal renal function, as defined by creatinine ⬎
`upper limit of normal, proteinuria 1⫹ or greater on 2 repeat
`determinations, urine protein:creatinine ratio ⱖ 1, or 24-hour
`urinary protein excretion of ⱖ300 mg. During the treatment period
`of this study, the only approved treatment for subfoveal choroidal
`neovascularization in patients with AMD was photodynamic ther-
`apy. Because photodynamic therapy was the standard of care, all
`patients who were eligible for photodynamic therapy were encour-
`aged to have photodynamic therapy and not enter the trial. Several
`of the patients had prior photodynamic therapy (5 subjects had
`multiple prior photodynamic therapy treatments, and 8 had at least
`1 prior photodynamic therapy) and still showed evidence of leak-
`age. All of these patients were eligible for photodynamic therapy
`and it was recommended, but they refused because they felt it was
`not benefiting them and they preferred to enter a trial rather than
`have additional photodynamic therapy. There were more of these
`patients (4/7 with prior photodynamic therapy) in the 0.3-mg/kg
`group than in the 1.0-mg/kg (2/7) and 3.0 mg/kg (2/5) groups due
`to random variation in a small study.
`
`Infusions of VEGF Trap or Placebo
`Patients were randomized 3:1 to receive VEGF Trap or an inactive
`saline infusion at each dose level. Masking of treatment assign-
`ment was the responsibility of the investigational pharmacist at
`each participating center. Individual patient doses were prepared
`by diluting the appropriate volume of VEGF Trap (25 mg/ml) with
`0.9% sodium chloride to yield a final concentration of 4 mg/ml.
`The volume of solution to be prepared was 25 to 150 ml, depend-
`ing on the patient’s dose and body weight. Patients randomized to
`receive placebo infusions received an infusion of a volume of
`normal saline (0.9% sodium chloride) equivalent to the volume of
`active drug solution that would be administered for the patient’s
`weight and dose level. The drug or placebo was infused over a
`period of no less than 1 hour by a registered nurse or physician’s
`assistant under the guidance of one of the investigators. In addi-
`tion, internists or anesthesiologists were also coinvestigators at all
`sites and helped to oversee the administration of the study drug and
`aid in the management of adverse events.
`
`Study Activities and Assessments
`Subjects were monitored closely for safety and tolerability using
`the following assessments and procedures: slit-lamp biomicros-
`copy, indirect ophthalmoscopy, tonometry, BCVA measurement,
`adverse event reporting, vital signs, physical examinations, serum
`electrolytes, creatinine, quantitative protein determination in 24-
`hour urine specimens, and measurement of serum neutralizing
`antibodies directed against VEGF Trap. Hypertension was graded
`based on the Common Terminology Criteria for Adverse Events.
`Grade 2 hypertension is defined as a recurrent or persistent (ⱖ24
`hours), or symptomatic increase in BP by ⬎20 mmHg (diastolic)
`or to ⬎150/100 if previously within normal limits; monotherapy
`may be indicated. Grade 3 hypertension is defined as a recurrent or
`persistent or symptomatic increase in BP by ⬎20 mmHg (dia-
`stolic) or to ⬎150/100 if previously within normal limits that
`requires treatment with more than one drug or more intensive
`therapy than previously. However, for purposes of this study,
`grade 3 hypertension that could be easily controlled with a com-
`monly used combination of 2 agents (e.g., a diuretic, an angioten-
`sin converting enzyme inhibitor, and/or a ␤-blocker) was not
`considered a dose-limiting toxicity (DLT). This modification of the
`criteria was made to reflect current standard clinical practice in the
`management of hypertension. Hypertension that met the criteria
`for a grade 3 increase and could not be controlled with a combi-
`
`
`
`
`Apotex Exhibit 1023
`Page 3 of 14
`
`

`

`Nguyen et al
`
`䡠 VEGF Trap and Choroidal Neovascularization
`
`nation of commonly used antihypertensives was considered grade
`3. Grade 4 or malignant hypertension is hypertension with life-
`threatening consequences (e.g., hypertensive crisis). A DLT was
`defined as any Common Terminology Criteria for Adverse Events
`grade 3 or 4 toxicity or grade 2 or 3 ocular toxicity. Any grade 1
`or 2 toxicity that resulted in dose reduction or discontinuation of
`study drug was considered a potential DLT to be reviewed by the
`Investigator and Study Director. Other DLTs were urinary protein
`excretion of ⬎2 g/24 hours or Common Terminology Criteria for
`Adverse Events grade 2 or greater potential immunotoxicity,
`including but not limited to allergic reaction/hypersensitivity
`(including drug fever), autoimmune reaction, vasculitis, ery-
`thema mutliforme, rash/desquamation, urticaria, and/or asymp-
`tomtic bronchospasm. This includes infusion-related hypersen-
`sitivity reactions, symptoms of which may include flushing,
`dyspnea, tachycardia, bronchospasm (symptomatic or asymptom-
`atic), hypotension, anxiety, myalgias, edema, and nausea.
`Stereoscopic color fundus photography and fluorescein angiog-
`raphy (FA) were performed at baseline and days 29, 71, and 99.
`Optical coherence tomography was performed at each study visit.
`Figure 1 (available at http://aaojournal.org) shows a flowchart of
`study activities. Patients in the first and second cohorts received all
`4 infusions and were monitored through all study visits. All of the
`patients in the 3.0-mg/kg dose group were withdrawn from study
`after the DLT was identified. The first enrolled patient received all
`4 infusions; the second patient received 2 infusions; and the third,
`fourth, and fifth patients each received 1 infusion.
`
`Reading Center
`The Retinal Imaging Research and Reading Center (RIRRC) at the
`Wilmer Eye Institute served as the reading center for fluorescein
`angiographic and OCT analyses. All images were evaluated by a
`grader who was masked with respect to treatment group. Interpre-
`tation of optical coherence tomograms is more difficult in patients
`with choroidal neovascularization than in patients with diabetic
`macular edema. In patients with choroidal neovascularization, the
`computer often misinterprets borders, and therefore, the computer-
`generated foveal thickness, on which eligibility is based, may be
`misleading. In some instances, baseline scans that had computer-
`generated values that were ⬎250 ␮m, thereby allowing patients to
`be entered, had lower values when read later by the reading center
`in masked fashion. The values obtained in the reading center are
`more accurate because the borders selected by the computer for
`measurements were scrutinized and, when incorrect, were reset
`manually to get a more accurate reading.
`
`Optical Coherence Tomography
`Optical coherence tomography was performed using StratusOCT
`(Carl Zeiss Meditec, Dublin, CA). The RIRRC provided detailed
`instruction in the protocol for image acquisition, and a represen-
`tative from the RIRRC visited each study site to certify compe-
`tence and compliance. Two standard protocols (6-mm fast macular
`thickness map and 6⫻6-mm crosshair) and 1 modified acquisition
`protocol (3-line 8-mm papillomacular axis scan) were used. The
`3-line 8-mm papillomacular axis scans utilized the disc as a
`landmark to ensure reproducible placement of scan lines at each
`visit; 1 line was at the superior margin, 1 was at the inferior
`margin, and 1 passed through the center of the disc. The 6⫻6-mm
`crosshair was a high-resolution scan used to follow morphological
`changes in the macula. The fast macular thickness map performed
`6 linear scans 6 mm in length centered on the patient fixation at
`equally spaced angular orientations in 1.96 seconds. Retinal thick-
`ness at any point was defined as the distance between outer and
`inner reflectivity bands of the OCT cross section. Foveal thickness
`
`(in micrometers; defined as the mean height of the neurosensory
`retina in a central 1-mm-diameter area) and total macular volume
`(in cubic millimeters) were computed automatically by the Stra-
`tusOCT software (version 4.0). Due to the advanced nature of the
`disease and extensive change in the RPE morphology, manual scan
`profiling was used to correct any artifacts produced by the auto-
`mated analysis algorithm.
`
`Fluorescein Angiography
`
`High-resolution digital FA was performed using an FF4 fundus
`camera (Carl Zeiss Meditec, Oberkochen, Germany) attached to an
`MRP (Boston, MA) capture station. A modified FA acquisition
`protocol was used for image acquisition, and compliance was
`monitored by a site visit. Digital mages were sent to the RIRRC
`and analyzed using EyeRoute Proview software (version 6.1, Anka
`Systems Inc., McLean, VA). Two independent
`investigators
`graded each FA for markers of disease activity (progression/
`regression), including blood and pigment epithelial detachment.
`Each fluorescein angiogram was analyzed further using advanced
`image segmentation techniques for edge detection to determine the
`lesion size and maximum area and extent of leakage.
`
`Statistical Methods
`
`All statistical analyses were performed in SAS (version 9.0, SAS
`Institute, Cary, NC). Changes in continuous measures were as-
`sessed using analysis of covariance with a main effects model that
`included baseline excess foveal thickness as a covariate and treat-
`ment effects. Due to the small numbers of patients in this study, the
`assumptions of the parametric model of homoscedasticity and
`normality were examined, and if the parametric assumptions were
`unwarranted, rank analogs were to be advanced. A rank analysis of
`covariance was then to be performed. In fact, due to the small
`number of subjects in this study, parametric assumptions were not
`met; therefore, the primary analysis method was the rank analysis
`of covariance with a main effects model.
`
`Results
`
`Patient Population
`
`Nine patients were enrolled in the first cohort (0.3-mg/kg VEGF
`Trap and placebo), 9 in the second (1.0-mg/kg and placebo), and
`8 in the third (3.0-mg/kg and placebo). Figure 1 (available at
`http://aaojournal.org) shows the patient disposition in the study.
`One patient in the second cohort who was thought initially to have
`occult choroidal neovascularization was determined by the reading
`center not
`to have it. Although safety data could have been
`obtained from this patient, because there was no possibility for the
`patient to receive any benefit, it was judged inappropriate to
`expose the patient to any possible risk; therefore, the patient was
`withdrawn from the study. In addition, when DLT was observed in
`a patient in the 3-mg/kg cohort, dosing was stopped for all patients
`in that cohort, and the study was terminated.
`Table 1 shows the demographics of the study subjects. The age
`and gender distributions were similar across all 4 groups. By
`chance, subjects who were randomized to receive placebo treat-
`ment had worse VA at baseline than patients randomized to
`receive VEGF Trap. Lesion sizes were similar across all groups,
`except for the 0.3 mg/kg, which had a smaller average lesion size.
`Each group had a combination of predominantly classic, occult,
`and minimally classic choroidal neovascularization lesions.
`
`1522.e4
`
`
`
`
`Apotex Exhibit 1023
`Page 4 of 14
`
`

`

`Ophthalmology Volume 113, Number 9, September 2006
`
`Table 1. Demographics and Disease History
`
`Age (yrs) [mean (range)]
`Gender (male:female)
`ETDRS letters read [mean (range)]
`Lesion size (DAs) (mean)
`Lesion type (Occult:classic:minimally classic)
`Foveal thickness (␮m) [mean (range)]
`
`Placebo
`(n ⴝ 6)
`
`76.7 (64–86)
`3:3
`27.8 (12–50)
`6
`2:3:1
`384.4 (300–483)
`
`0.3 mg/kg
`(n ⴝ 7)
`
`76.3 (58–81)
`3:4
`47.9 (24–69)
`3
`4:2:1
`299.6 (238–340)
`
`1 mg/kg
`(n ⴝ 7)
`
`79.6 (68–88)
`2:5
`49.9 (16–64)
`6.5
`1:1:3
`348.8 (221–680)
`
`3 mg/kg
`(n ⴝ 5)
`
`73.8 (69–81)
`3:2
`47.8 (8–72)
`7
`1:2:2
`414.4 (289–563)
`
`DA ⫽ disc area; ETDRS ⫽ Early Treatment Diabetic Retinopathy Study.
`
`Preliminary Safety Assessment
`The most common adverse events were headache, hypertension,
`proteinuria, and hoarseness (Table 2). Hypertension, proteinuria,
`and hoarseness were expected adverse events, because they are
`class effects of systemic VEGF antagonists and had been seen in
`prior studies of VEGF Trap in patients with advanced malignan-
`cies (Proc ASCO abstract 3009, 2004). Adverse events were dose
`related and were most common in the 1.0- and 3.0-mg/kg cohorts
`(Table 2), but they were mild and easily managed in the 1.0-mg/kg
`group.
`VEGF Trap administration was associated with a dose-
`dependent increase in mean BP. The highest readings generally
`were recorded 2 weeks after the first dose. Table 3 shows the
`change from baseline to day 15 (2 weeks after dose) in mean
`systolic and diastolic BPs. The increase in diastolic pressure was
`statistically significant at all dose levels studied. By 4 weeks after
`the first dose, BP generally returned to baseline (data not shown),
`except in one patient whose clinical course is described below. All
`patients with BP elevations were treated successfully for their
`hypertension during the repeated-dosing portion of the study.
`Increases in mean BP were not noted past day 29.
`Severe hypertension and proteinuria occurred only in patients
`who received 3 mg/kg (Table 4). In one patient in the 3-mg/kg
`cohort, proteinuria reached a level predefined as dose limiting, and
`in another patient, hypertension was dose limiting (Table 5 [avail-
`able at http://aaojournal.org]). The patient with severe hyperten-
`sion was slow to respond to a change in antihypertensive regimen
`and developed congestive heart failure with pulmonary edema
`requiring hospitalization. This resolved with diuresis and more
`aggressive management of hypertension. Due to these DLTs, no
`additional VEGF Trap was administered to patients in the 3.0-
`mg/kg cohort, but follow-up was continued to monitor safety. All
`adverse events resolved after VEGF Trap was discontinued. As a
`result of the safety analysis, 1.0 mg/kg was determined to be the
`
`maximum tolerated intravascular dose of VEGF Trap in this pa-
`tient population.
`
`Foveal Thickness by Optical Coherence
`Tomography
`
`The prespecified primary measure of bioactivity was the effect on
`excess foveal thickness assessed by OCT. Normal foveal thickness
`is 179⫾17 ␮m19; therefore, excess foveal thickness is measured as
`foveal thickness ⫺ 179 ␮m. Figure 2A presents sequential foveal
`thickness values for each patient in the 1.0-mg/kg cohort. The
`normal range of 145 to 213 ␮m, 2 standard deviations (SDs) above
`and below 179 ␮m, is shaded. Three of 5 patients who received
`1.0-mg/kg VEGF Trap had reduction of foveal thickness into the
`normal range, and the other 2 were close to the normal range,
`whereas the 2 patients treated with the placebo had little or no
`reduction in foveal thickness. It is notable that every patient treated
`with 1.0-mg/kg VEGF Trap had reduction in foveal thickness 2
`weeks after the first infusion and an increase over the next 2 weeks,
`and then most had reduction after the second, third, and fourth
`infusions.
`Figure 2B illustrates the median percentage change in excess
`foveal thickness. Drug infusions are shown by arrows located at
`the top of the figure (days 1, 29, 43, and 57). Eight days after the
`initial infusion, all 3 cohorts treated with VEGF Trap showed a
`decrease in excess foveal thickness, but the effect with 0.3 mg/kg
`was marginal and lost at subsequent time points. In patients treated
`with 1.0-mg/kg VEGF Trap, there was a persistent effect at 2
`weeks with elimination of 70% of excess foveal thickness, but by
`1 month, the effect was reduced; there was a 40% decrease in
`excess foveal thickness. After subsequent infusions of 1.0-mg/kg
`VEGF Trap, there was again reduction in excess foveal thickness,
`and at day 71, 2 weeks after the fourth infusion, excess foveal
`thickness was almost eliminated (80% reduction). Although only a
`
`Table 2. Adverse Events
`
`Placebo
`(n ⴝ 6)
`
`1 (16.7)
`1 (16.7)
`1 (16.7)
`
`0.3 mg/kg
`(n ⴝ 7)
`
`1 (14.3)
`
`1 (14.3)
`1 (20.0)
`
`1.0 mg/kg
`(n ⴝ 7)
`
`3.0 mg/kg
`(n ⴝ 5)
`
`All Doses
`(n ⴝ 19)
`
`3 (42.9)
`3 (42.9)
`3 (42.9)
`1 (14.3)
`1 (14.3)
`2 (28.6)
`
`4 (80.0)
`3 (60.0)
`3 (60.0)
`3 (60.0)
`1 (20.0)
`
`3 (60)
`
`8 (42.1)
`6 (31.6)
`6 (31.6)
`5 (26.3)
`3 (15.8)
`3 (15.8)
`3 (15.8)
`
`Headache
`Hypertension
`Proteinuria
`Hoarseness
`Arthralgia
`Cough
`Aggravation of arthritis pain
`
`n (% of subjects at dose level).
`
`1522.e5
`
`
`
`
`Apotex Exhibit 1023
`Page 5 of 14
`
`

`

`Nguyen et al
`
`䡠 VEGF Trap and Choroidal Neovascularization
`
`Table 3. Mean (⫾ Standard Deviation) Change from Baseline
`in Systolic and Diastolic Blood Pressures 15 Days after Initial
`Dose of Vascular Endothelial Growth Factor Trap
`
`Systolic
`(mmHg Change)
`⫺1.8⫾11.2
`⫹3.0⫾12.7
`⫹8.2⫾27.4
`⫹26.8⫾17*
`
`Diastolic
`(mmHg Change)
`⫹3.4⫾6.7
`⫹11.0⫾4.0*
`⫹14.7⫾6.4*
`⫹20.4⫾10.1†
`
`Dose Group
`Placebo (n ⫽ 6)
`0.3 mg/kg (n ⫽ 7)
`1.0 mg/kg (n ⫽ 7)
`3.0 mg/kg (n ⫽ 5)
`
`*P⬍0.05.
`†P⬍0.01.
`
`limited amount of data are available for patients treated with
`3.0-mg/kg VEGF Trap, the effect after the first infusion appeared
`to be maintained for at least 1 month.
`
`Macular Volume by Optical Coherence
`Tomography
`Total macular volume is a measure of retinal thickness over a
`broader area than that indicated by foveal thickness. Previous
`studies have demonstrated that 6.47 mm3 is the average macular
`volume in patients with no macular disease (Shah, Nguyen, Cam-
`pochiaro, unpublished data). Three of the 5 patients treated with
`1.0-mg/kg VEGF Trap had baseline total macular volume mea-
`surements more than 2 SDs above 6.47 mm3. All 3 patients had
`substantial reduction in total macular volume, and on day 71, 2
`weeks after the last VEGF Trap infusion, all had total macular
`volume measurements very close to 6.47 mm3, indicating resolu-
`tion of essentially all excess volume (Fig 3A [available at http://
`aaojournal.org]). By 6 weeks after the last infusion (day 99), total
`macular volume had begun to increase again. Two patients had
`normal total macular volume at baseline, although their foveal
`thickness was elevated; a possible explanation is that these patients
`had some areas of retinal thinning due to scarring and atrophy that
`counter areas of thickening when measurements are taken over a
`broad area. These patients also showed reduction in total macular
`volume that rebounded 6 weeks after completion of VEGF Trap
`infusions. The change from baseline in excess total macular vol-
`ume mirrored the change from baseline in excess foveal thickness,
`with si