AMERICAN ACADEMY*®
`OF OPHTHALMOLOGY
`The Eye M.D, Association
`
`Safety and Efficacy of Conbercept in
`Neovascular Age-Related Macular
`Degeneration
`Results from a 12-Month Randomized Phase 2 Study:
`AURORA Study
`
`
`
`Xiaoxin Li, MD, PhD,' Gezhi Xu, MD,” Yusheng Wang, MD,* Xun Xu, MD,* Xiaoling Liu, MD,”
`Shibo Tang, MD,° Feng Zhang, MD,’ Junjun Zhang, MD,° Luosheng Tang, MD,’ Quan Wu, PhD,'°
`Delun Luo, BS,'° Xiao Ke, BS,'° for the AURORA Study Group
`
`Purpose: To assessthe safety and efficacy of multiple injections of 0.5 and 2.0 mg conberceptusing variable
`dosing regimensin patients with neovascular age-related macular degeneration (AMD).
`Design: Randomized, double-masked, multicenter, controlled-dose, and interval-ranging phase 2 clinical
`trial divided into a 3-month loading phase followed by a maintenance phase.
`Participants: Patients with choroidal neovascularization secondary to AMD with lesion sizes of 12 disc areas
`or less and a best-corrected visual acuity (BCVA) letter score of between 73 and 24 were enrolled.
`Methods: Patients were randomized 1:1 to receive either 0.5 or 2.0 mg intravitreal conbercept for 3
`consecutive monthly does. After the third dose, each group was reassigned randomly again to monthly (Q1M
`group) or as-needed(pro re nata [PRN] group) treatment without changing the drug assignment.
`Main Outcome Measures: The primary end point was the mean change in BCVA from baseline to month 3,
`with secondary end points being the mean change in BCVA, mean changein central retinal thickness (CRT), and
`safety at month 12.
`Results: We enrolled 122 patients. At the primary end point at month 3, mean improvements in BCVA from
`baseline in the 0.5- and 2.0-mg groups were 8.97 and 10.43 letters, respectively. At month 12, mean improve-
`ments in BCVA from baseline were 14.31, 9.31, 12.42, and 15.43 letters for the 0.5-mg PRN, 0.5-mg Q1M, 2.0-mg
`PRN, and 2.0-mg Q1M regimens, respectively. At month 12, mean reductions in CRT in the 4 regimens were
`119.8, 129.7, 152.1, and 170.8 |im, respectively. There were no significant differences for the pairwise com-
`parisons betweenall study groups. The difference in the numberof injections between the 2 PRN groups wasnot
`statistically significant. Treatment with conbercept generally was safe and well tolerated.
`Conclusions: The significant gains in BCVA at 3 months were the sameor better at 12 monthsin all con-
`bercept dosing groups of neovascular AMD patients. During the 12 months, repeatedintravitreal injections of
`conbercept were well tolerated in these patients. Future clinical trials are required to confirm its long-term efficacy
`and safety. Ophthalmology 2014;121:1740-1747 © 2014 by the American Academy of Ophthalmology. This is an
`open accessarticle under the CC BY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/3.0/).
`
`Age-related macular degeneration (AMD) is a progressive
`disease of the macula and the leading cause ofirreversible
`blindness in industrialized countries.’ Althoughit hasnotyet
`become the leading cause of blindness among the Chinese
`population, the prevalence of AMDis rising gradually as
`the population ages
`and the
`socioeconomic
`situation
`improves.” An epidemiologic investigation showed that
`15.5% of the included Shanghai residents (>50 years of
`age) had AMD and 11.9% of them had neovascular
`(exudative) AMD.* Neovascular AMD is characterized by
`the growth of abnormal new blood vessels under the retinal
`pigment epithelium, under the retina, or within the retina.
`Whenneovascularization arises from the choroid, these new
`blood vessels are referred to as choroidal neovascularization
`
`(CNV).“ The pathophysiologic features of neovascular
`AMDarenotfully understood, but it is known that vascular
`endothelial growth factor (VEGF) plays an importantrole in
`the proliferation and maintenance of this neovascularization.
`This
`fact has
`led to the development of
`therapeutic
`strategies to inhibit VEGFfor the treatment of neovascular
`AMD.”
`Between 2004 and 2006, three anti-VEGF drugs were
`introduced to ophthalmology after either receiving regula-
`tory approval for the treatment of AMD orbeing used in an
`off-label manner. They exhibit important differences in their
`sites of activity, formulation methods, binding affinities, and
`biologic activities. Pegaptanib (Macugen; Eyetech Pharma-
`ceuticals, Lexington, MA)is a ribonucleic acid aptamerthat
`
`1740
`
`© 2014 by the American Academy of Ophthalmology
`This is an open accessarticle under the CC BY-NC-ND license
`RegeneronExhibit2027“ "°1
`Page 01 of 08
`
`http://dx.doi.org/10.1016/j.ophtha.2014.03.026
`ISSN 0161-6420/14
`Miylan v. Regeneron, IPR2021-00880
`U.S. Pat. 9,669,069, Exhibit 2027
`
`
`
`Regeneron Exhibit 2027
`Page 01 of 08
`
`

`

`Li et al
`
`+ Conbercept for Neovascular AMD
`
`blocks the main pathologic isoform of VEGF (known as
`efficacy of different doses and different dosing regimens were
`compared after repeated intravitreal injections of conbercept. The
`VEGF165) andlarger isoforms of VEGF byattaching to its
`heparin binding domain,° whereas ranibizumab (Lucentis;
`primary end point was assessed at month 3, and the results of the
`maintenance phase were assessed at month 12. The major eligi-
`
`
`
`
`Genentech, San_Francisco, CA)Inc., South and
`bility criteria included age 50 years or older, the presence in the
`bevacizumab (Avastin; Genentech and Roche, Basel,
`study eye (1 eye per patient) of untreated active subfoveal or
`Switzerland)
`are derived from a murine monoclonal
`Juxtafoveal CNV secondary to AMD, lesion size 12 disc areas or
`antibody against WEGF-A;
`ranibizumab is an_affinity-
`less in either eye, and BCVAletter scores in the study eye between
`73 and 24. The BCVA score was based on the numberofletters
`matured, humanized, monoclonal antigen binding fragment
`from the antibody and bevacizumab is a full-length,
`read correctly on the Early Treatment Diabetic Retinopathy Study
`humanized, monoclonal antibody directed against VEGF-
`visual acuity chart when assessed at a starting distance of 4 m. An
`A. Both drugs function by blocking the same receptor
`Early Treatment Diabetic Retinopathy Study visual acuity score of
`binding domains of all VEGF-A isoforms.’ In November
`73 to 24letters is approximately 20/40 to 20/320in Snellen visual
`acuity. An increase in the BCVA letter score indicates improve-
`2011, aflibercept (Eylea; Regeneron, Tarrytown, NY; and
`ment
`in visual acuity. Patients were excluded if any of the
`Bayer, Leverkusen, Germany) was approved by the US
`following were present: significant subfoveal atrophy or scarring;
`Food and Drug Administration. This
`soluble decoy
`presence of other causes of CNV in either eye; history of previous
`receptor is produced by combining all-human DNA se-
`AMD drug treatment (such as anti-VEGF drugs and steroids);
`quences of the second binding domain of human VEGF
`previous laser therapy or other ocular operation, or both,
`in the
`receptor (VEGFR)-1 to the third binding domain of human
`study eye, such as macular translocation surgery, cataract surgery,
`VEGFR-2, which is then combined with the Fc region of
`vitrectomy surgery, glaucoma filtering operation, verteporfin
`human immunoglobulin G-1.° Aflibercept binds to all
`photodynamic therapy, subfoveal focal laser photocoagulation, and
`VEGF-A and VEGF-Bisoforms, as well as to the highly
`transpupillary thermotherapy;
`active ocular
`inflammation or
`infection; uncontrolled diabetes mellitus; uncontrolled hyperten-
`related placental growth factor.
`sion; history of cerebrovascular accident or myocardial infarction
`(KH902; Chengdu
`Similar
`to aflibercept, conbercept
`within 6 months; renal failure requiring dialysis or renal transplant;
`KanghongBiotech Co., Ltd., Sichuan, China) consists of the
`pregnancy or lactation; or history of allergy to fluorescein or
`VEGF binding domains of the human VEGFR-1 and
`povidone
`iodine.
`The
`trial was
`registered
`at www.
`VEGFR-2 combined with the Fe portion of the human
`clinicaltrial.gov under the identifier NCT 01157715.
`immunoglobulin G-1. In addition to having high affinity for
`all isoforms of VEGF-A,it also binds to placental growth
`factor and VEGF-B. The structural difference between
`conbercept and aflibercept is that conbercept also contains
`the fourth binding domain of VEGFR-2. This fourth domain
`is essential for receptor dimerization and enhances the as-
`sociation rate of VEGF to the receptor.”'° Because this
`domain of VEGFR-2 has a lower isoelectric point,
`the
`addition of this domain to KH902 decreases the positive
`charge of the molecule and results in decreased adhesion to
`the extracellular matrix. Preclinical studies have demon-
`strated that conbercept showsstrong antiangiogenetic effects
`by binding with high affinity and neutralizing VEGF-A,all
`its isoforms, and placental growth factor.''
`Intravitreal administration of conbercept has been shown
`to successfully prevent lesion growth and leakage of CNV
`in a nonhuman primate model.'':!* A phase | study also
`demonstrated that conbercept resulted in improvements in
`best-corrected visual acuity (BCVA), reduction in central
`retinal thickness (CRT), and a decrease in the area of CNV
`in patients with neovascular AMD." The present study was
`designed to investigate the safety and efficacy of intravitreal
`injections of conbercept in patients with CNV secondary to
`AMD.
`
`Intervention
`
`Eligible patients were randomized 1:1 to 0.5- or 2.0-mg treatment
`groups.
`Initially, all patients received monthly intravitreal
`in-
`jections of conbercept for a total of 3 injections. After the 3-month
`loading phase, patients were reassigned randomly to monthly
`(Q1M group) or as-needed treatments (pro re nata [PRN] group)
`with the same dose of conbercept given during the loading phase.
`Patients randomized to the monthly regimen were treated
`monthly during the maintenance phase. Patients randomized to the
`PRN regimen were notre-treated unless any of the following was
`present in the study eye: a more than 100-{1m increase in CRT
`compared with the lowest previous measurement; a loss of 5 or
`more BCVAletters compared with the best previous measurement;
`new,recurrent, or persistent subretinal or intraretinal fluid based on
`the review ofall the optical coherence tomography (OCT) scans;
`new onsetof classic neovascularization; new or persistent leakage
`on fluorescein angiography (FA); or new macular hemorrhage or
`hemorrhagic area of more than 50%of the disc area. Decisions
`about re-treatment were made on the basis of the investigator’s
`evaluation of the BCVA, ophthalmic examination results, and
`images from OCT, FA, and fundus photography (FP). The inves-
`tigator was masked to the assignment of dose in the PRN arms.
`Rescue therapy with another treatment was not offered as part of
`this study, so if a patient elected to receive any other therapy for
`their neovascular AMD,then they were askedto exit the study. The
`only approved anti-VEGF therapy in China is ranibizumab, and
`ranibizumab was not approved in China until 2012, which occurred
`well after the start of this study in 2010.
`The study was conducted in accordance with the Declaration of
`Helsinki and its subsequent amendments, China good clinical
`practice regulations, and applicable institutional
`regulatory re-
`quirements. Before the initiation of the study, relevant institutional
`review boards and ethics committees from the respective study
`centers approved the research protocol and its amendments. All
`patients provided written informed consent for study participation.
`
`174]
`
`Methods
`
`Study Design
`
`randomized, double-
`The AURORA study was a 12-month,
`masked, controlled-dose, and interval-ranging phase 2 clinical
`trial and was designed as a superiority trial to assess the safety and
`efficacy of different dosing regimens of conbercept in patients with
`CNV secondary to AMD. At 9 sites in China,
`the safety and
`
`Regeneron Exhibit 2027
`Page 02 of 08
`
`Regeneron Exhibit 2027
`Page 02 of 08
`
`

`

`Ophthalmology Volume 121, Number 9, September 2014
`
`Assessments
`
`Outcomes
`
`All patients were evaluated monthly. Evaluations included visual
`function, ocular assessments,
`adverse
`events using BCVA
`measured with the Early Treatment Diabetic Retinopathy Study
`chart (4-m starting distance), intraocular pressure measurements,
`slit-lamp examinations,
`and imaging with FP, OCT, FA,
`and indocyanine green angiography (ICGA). Fundus photography
`and OCT imaging were performed at every visit, whereas FA and
`ICGA were performed only at baseline and at months 3, 8, and
`12. Optical coherence tomography was performed either with the
`Stratus OCT instrument (Carl Zeiss Meditec, Dublin, CA) or the
`Heidelberg Spectralis spectral-domain OCT instrument (Heidel-
`berg Engineering, Heidelberg, Germany). The same type of OCT
`instrument used at baseline was used throughout the study. When
`the Stratus OCT was used,
`the following scan patterns were
`performed on both eyes and were centered on the fovea: two 7-
`mm posterior pole custom scans positioned 5° below horizontal
`from the temporal edge of the optic nerve toward the fovea (512
`A-scans per B-scan), one 3-mm high-resolution cross-hair scan
`(512 A-scans per B-scan), one 6-mm high-resolution cross-hair
`linear scan (512 A-scans per B-scan), and 2 fast macular thickness
`map scans consisting of 6 radial linear scans (128 A-scans per
`B-scan). When the Heidelberg Spectralis spectral-domain OCT
`was used, the following scan patterns were performed on both
`eyes and were centered on the fovea: a single 30° horizontal
`section scan with an automatic real-time setting of 15 (1536
`A-scans per B-scan) and a volume scan over a 20°x20° area
`consisting of 49 B-scans (512 A-scans per B-scan), with each B-
`scan separated by 120 um, and an automatic real-time setting of
`15. Either the Topcon TRC.50-DX or the Heidelberg HRA2 were
`used to perform FA and ICGA. Fundus photography was per-
`formed using the Topcon TRC.50-DX, Topcon TRC-50EX, and
`Zeiss FF 450 plus. The OCT, FA, ICGA, and FP images were
`graded at a central
`reading center
`(the Digital Angiography
`Reading Center, New York, NY). Adverse events (AEs) were
`recorded at each visit as well. Study visits were scheduled every
`
`3047 days.
`
`The primary efficacy outcome was the mean change in BCVA
`score from baseline at month 3. Secondary outcomes at month 12
`were the mean changes of BCVA score from baseline over time,
`the incidence rates of AEs over time, the mean changes in CRT on
`OCT imaging over time, the changes in leakage area on FA im-
`aging, and the mean numberofinjections over time.
`
`Statistical Methods
`
`The full analysis dataset with all the patients who completed the
`month 12 visit was the dataset used for the primary efficacy
`analysis. Mean changes in BCVA from baseline at months 3 and
`12 were assessed using the paired f test or rank-sum test with 95%
`confidenceintervals. The chi-square test or Fisher test was used for
`the proportions of patients who gained more than 0letters, gained
`at least 15 letters, and gained atleast 30 letters or lost fewer than 15
`letters. Other secondary end points, as well as demographic data at
`baseline, were evaluated using summarystatistics.
`The safety analysis set with all the patients who participated in
`the study was used for all safety and tolerability assessments. All
`the AEs,
`treatment-related AEs,
`incidence of AEs, and serious
`AEs (SAEs) were compared between groups using the chi-square
`test or Fisher exact method. All statistical tests were 2-sided. A
`P value less than 0.05 was consideredstatistically significant. All
`the above analyses were performed using SAS software version 9.1
`(SASInc., Cary, NC). Adverse events were coded with the Med-
`ical Dictionary for Regulatory Activities (MedDRA 14.1; the In-
`ternational Federation of Pharmaceutical Manufacturers
`and
`Associations [IFPMA], Geneva, Switzerland).
`
`Results
`
`Characteristics of Patients
`
`Between July 2010 and July 2012, 122 patients (0.5-mg group, n =
`60; 2.0-mg group, n = 62) were randomized. Onepatient withdrew
`
`fcc
`| Baseline
`
`1
`
`0.5 mg group Q1M
`
`Enrollment
`(n=122)
`|
`1st Randomization J
`
`| Withdraw
`n=4
`
`
`(n=60) -----4>
`
`
`%
`
`0.5 mg group QiM
`CLT
`(n=56) group
`i End of loading y_ .
`Y= phase _mi i i a a>
`nd
`)
`( Randomization |
`0.5 mg Q1M group
`0.5 mg PRN group
`(n=29)
`(n=26)
`-- >| Withdraw
`n=2
`(n=24)
`0.5 mg PRN group
`2.0 mg Q1M group
`(n=29)
`
`(n=62)
`-~---» Withdraw |
`ke
`n=4
`
`-
`
` 2.0 mg group QiM
`
`
`
`
`won|
`ond
`|
`__ Randomization |
`2.0 mg Q1M group
`2.0 mg PRN group
`(n=30)
`(n=26)
`
`— (n=58) ee
`Withdraw
`Withdraw |
`n=1
`n=2
`J
`
`-- > Withdraw |
`\
`n=3
`)
`
`J
`
`—->{ Withdraw )
`L
`n=1
`
`(~ Endof~ “1
`| maintenance |—
`\_ _pbase_ _/
`
`_|
`
`0.5 mg Q1M group
`(n=26)
`
`2.0 mg PRN group
`
`(n=26)
`
`Figure 1. Diagramofparticipant flow in the AURORAstudy. Before the second randomization, 5 (8.3%) and 6 (9.7%) patients withdrew from the 0.5-mg
`monthly (QIM) and 2.0-mg Q1M groups,respectively. At the end of maintenance phase, 3 (10.3%), 2 (7.7%), 1 (3.3%), and 0 (0%) patients prematurely
`discontinued the study in the 0.5-mg QIM, 0.5-mg as-needed (pro re nata [PRN]), 2.0-mg QIM, and 2.0-mg PRN groups,respectively. A total of 102
`patients were included in the finalanalysis.
`
`1742
`
`Regeneron Exhibit 2027
`Page 03 of 08
`
`Regeneron Exhibit 2027
`Page 03 of 08
`
`

`

`Li et al
`
`+ Conbercept for Neovascular AMD
`
`Table 1. Baseline Characteristics of the Study Population
`
`2.0-mg Group
`0.5-mg Group
`
`Characteristics
`As Needed (n = 26)
`Monthly (n = 29)
`As Needed (n = 26)
`Monthly (n = 30)
`
`63.5347.55
`
`64,508.89
`
`69.6648.26
`
`66.0849.27
`
`Age (yrs)
`Sex, no. (%)
`Male
`Female
`Study eye, no. (%)
`Right eye
`Left eye
`BCVA(letters)
`CRT (im)
`CNVtype, no. (%)
`Occult
`Classic
`Predominantclassic
`CNV area (mm”)
`Fluorescein leakage, no. (%)
`Yes
`No
`Leakage area (mm)
`
`13 (50.0)
`13 (50.0)
`
`14 (53.9)
`12 (46.2)
`46.58414.54
`
`291.544+183.35
`
`19 (65.5)
`10 (34.5)
`
`13 (44.8)
`16 (55.2)
`50.79+12.87
`
`310.90+138.45
`
`5 (19.2)
`4 (15.4)
`17 (65.4)
`8.0748.07
`
`26 (100.0)
`0 (0.0)
`8.3047.87
`
`
`
`12 (41.4)
`6 (20.7)
`11 (37.9)
`8.40+6.14
`
`29 (100.0)
`0 (0.0)
`9.3146.28
`
`
`
`16 (61.5)
`10 (38.5)
`
`15 (57.7)
`L1 (42.3)
`
`47.62413.73
`330.364+121.24
`
`10 (38.5)
`5 (19.2)
`11 (42.3)
`
`9.75+6.50
`
`26 (100.0)
`0 (0.0)
`10.94+7.02
`
`22 (73.3)
`8 (26.7)
`
`17 (56.7)
`13 (43.3)
`48.87414.66
`
`335.50+152.39
`
`7 (23.3)
`5 (16.7)
`18 (60.0)
`7.7446.91
`
`30 (100.0)
`0 (0.0)
`8.2646.74
`
`
`
`BCVA = best-corrected visual acuity; CRT = central retinal thickness; CNV = choroidal neovascularization.
`Data are mean + standard deviation unless otherwise indicated.
`
`consent after the randomization and 1 patient was found to have
`angioid streaks, which was a condition among the exclusion
`criteria, leaving 120 patients who were treated and included in our
`analyses. Of
`these, 114 patients completed treatment
`in the
`3-month loading phase. Reasons for withdrawal included an SAE
`(n = 1, reduced visual acuity score from 40 to 5),
`investigator
`decisions (n = 2), protocol deviation (n = 1), an inability to attend
`visits (n = 1), and a subject’s request (n = 1). Before the second
`randomization, 3 patients exited the study. One withdrew consent
`and 2 had ocular AEs that included ocular inflammation and vit-
`reous opacities, leaving 111 patients to continue treatment in the
`maintenance phase. Overall, 105 patients (86.1%) completed the
`12-month study period (0.5-mg PRN, n = 24; 0.5-mg QIM, n =
`26; 2.0-mg PRN, n = 26; 2.0-mg QIM, n = 29; Fig 1). The
`reasons why the 6 patients exited before the final month 12 visit
`included SAEs (n = 3;
`including 1 case of suspected drug-
`induced hepatitis,
`1 case of hepatitis B, and 1 hepatic tumor), an
`investigator decision (n = 1; AMD progressin the fellow eye), an
`inability to attend visits (n = 1), and a subject’s request (n = 1).
`During the entire study, 10 patients were deemed ineligible
`because of protocol deviations. Eight of them failed to meet the
`study eye inclusioncriteria, and the other 2 did not meet nonocular
`inclusion criteria. Overall,
`the randomized groups were well
`balanced with respect to baseline demographics and study eye
`characteristics (Table 1).
`
`Efficacy
`
`Treatment with conbercept produced significant improvements in
`BCVAinall treatment groups at both month 3 (the primary end
`point) and month 12 (Table 2). Most of the improvement occurred
`during the loading phase inthe first 3 months. The mean changes
`
`in BCVAfrom baseline at month 3 were 8.97-13.08letters for the
`
`0.5-mg group (P<0.0001) and 10.43+10.65 letters for the 2.0-mg
`group (P<0.0001). Furthermore, these improvements were main-
`tained or increased during the study. At month 12, mean changesin
`BCVA were 14.31+17.07 letters (0.5-mg PRN; P = 0.0002),
`
`
`9.31+10.98 letters (0.5-mg QIM; P<0.0001), 12.42+16.39letters
`(2.0-mg PRN; P = 0.0007), and 15.43+14.70 letters (2.0-mg
`QIM; P<0.0001) compared with baseline (Fig 2). The visual
`outcomes from the 2 dosing regimens were compared along with
`visual outcomes
`from all
`the
`study groups. No significant
`differences were observed between the dosing regimens and all
`study groups using pairwise comparisons (P>0.05).
`At month 12, the proportions of patients gaining 15 letters or
`more were 50.0%, 31.0%, 42.3%, and 46.7% for the 0.5-mg PRN,
`0.5-mg QIM,2.0-mg PRN, and 2.0-mg Q1M groups, respectively.
`At month 12, the proportions of eyes losing fewer than 15 letters
`were 100.0%, 96.55%, 96.15%, and 100.0%, respectively (Fig 3).
`Improvements
`in BCVA with conbercept
`treatment were
`associated with a decrease in CRT measured with OCT imaging.
`
`Table 2. Best-Corrected Visual Acuity Outcomes at Months 3 and 12
`
`
`
`Time Point 2.0-mg Group 0.5-mg Group
`
`Month 3
`BCVA (letters)
`Change from BSL (letters)
`Month 12
`BCVA(letters)
`Change from BSL(letters)
`
`58.394+17.30
`59.46+16.13
`
`8.97413.08
`10.434+10.65
`As Needed (n = 26) Monthly (n = 29) Total (n = 55) As Needed (n = 26) Monthly (n = 30) Total (n = 56)
`60,.10+17.52
`60.884 17.42
`59,92+18.82
`60.04+18.54
`64.30+16.37
`62.644+17.12
`
`
`
`
`9.31+10.98
`14.31417.07
`10.494+15,99
`12.424+16.39
`15.43+14.70
`13.61414.97
`
`BCVA = best-corrected visual acuity; BSL = baseline.
`Data are mean + standard deviation.
`
`Regeneron Exhibit 2027
`Page 04 of 08
`
`1743
`
`Regeneron Exhibit 2027
`Page 04 of 08
`
`

`

`Ophthalmology Volume 121, Number 9, September 2014
`
`iE
`
`= 2
`
`0
`
`-30
`
`-180
`
`& 6=
`
`$
`
`
`
`3
`60
`a
`
`-90
`5
`&O 120 -
`£@
`2 -150 -
`
`—-0.5mg PRN
`<0.5mg Q1M
`=t2.0me PRN
`==2.0mg Q1M
`
`
`
`oOo
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`Month
`
`7
`
`8
`
`9
`
`10 1 2
`
`*P<0.001
`
`Figure 4. The mean changein centralretinal thickness (CRT)frombaseline
`overtime using the 4 dosing regimens through 12 months. The CRT reduced
`rapidly duringthe first 3-month loading phase and then continued todecrease
`through month 12. PRN = prore nata(as needed); QIM = monthly.
`
`inflammation. Most AEs were reported as mild or moderate and
`disappeared with or without treatment. During the entire study
`period, 39 patients (66.1%) in the 0.5-mg group reported AEs,
`which included 4 (6.78%) related to the study drug, 11 (18.64%)
`associated with intravitreal injections, and 7 (11.86%) SAEs; 45
`patients (73.77%)
`in the 2.0-mg group reported AEs, which
`included 4 (6.56%) related to the study drug, 17 (27.87%) asso-
`ciated with intravitreal injection, and 3 (4.92%) SAEs.
`During the maintenance phase,
`the incidence rates of ocular
`AEsin the study eyes were 23.1%, 20.7%, 27.0%, and 30.0% for
`the 0.5-mg PRN, 0.5-mg QIM, 2.0-mg PRN, and 2.0-mg QIM
`groups, respectively. The group with the highest exposure, the 2.0-
`mg QIM group, also had the highest rate of AEs. However,
`because of the limited sample size,
`this phase 2 study was not
`powered adequately to assess the significance of these differences
`in AEs amongthe treatment groups. The SAEsaffecting study eyes
`were uncommonin all treatment groups. One patient in the 0.5-mg
`PRN group received cataract extraction and intraocular lens im-
`plantation using phacoemulsification because of cataract progres-
`sion with reduction in BCVA (compared with baseline) during the
`study. The patient recovered well after surgery and did notexit the
`study. One patient in the 2.0-mg Q1M group was hospitalized after
`the last injection because of pain in the study eye, a decrease of 7
`letters in BCVA (decreased by 65 letters compared with baseline),
`foreign body sensation, and vitreous opacity. This patient under-
`went a tap for presumed endophthalmitis, and although the bac-
`terial culture results were negative,
`this patient was diagnosed
`clinically with infectious endophthalmitis and received antibiotic
`therapy. After the antibiotic therapy, the symptoms of inflamma-
`tion dissipated, with a concomitant gradual
`improvement
`in
`BCVA.By the last study visit, the BCVA was restored to 70let-
`ters, and the comea and lens were clear. The investigators judged
`that both SAEs might have been related to treatment. In addition, 2
`patients experienced visual acuity decreases of more than 30letters
`(compared with the last assessment of BCVA before the most
`recent treatment). One case occurred in the nonstudy eye. The other
`one occurred in the study eye during the loading phase with the
`0.5-mg dose, and the investigator thought it was in the patient’s
`best interest to exit the study.
`No systematic (nonocular) AE was judged by the investigators
`to be related to the study drug or to the study procedure. No events
`described by the Antiplatelet Trialists’ Collaboration occurred
`during the study. There were no cardiovascular or cerebrovascular
`events such as heart failure, stroke, or arterial thrombosis. There
`were no apparent allergic reactions, and there were no deaths
`during the study period. All SAEs, the frequent study drug—related
`AEs, and the study procedure—related AEs are summarized in
`Table 4.
`
`
`
`
`
`MeanchangeinBCVAfromBaseline
`
`(Letters)
`
`Month
`
`==0.5mg PRN
`--0.5mg Q1M
`te2.0mg PRN
`<2.0mg Q1M
`
`=P <0.001
`
`Figure 2. The mean change in best-corrected visual acuity (BCVA) from
`baseline over time in patients in the 4 dosing regimen treatment groups
`through 12 months. PRN = prore nata (as needed); QIM = monthly.
`
`The CRT decrease observed at month 3 continued to decrease
`
`through month 12. By month 12, the mean CRT measurements had
`decreased by 116.0£194.84 [um (0.5-mg PRN; P = 0.0056),
`131.6+180.42 tum (0.5-mg QIM; P = 0.0005), 157.84183.98 um
`(2.0-mg PRN; P = 0.0003), and 168.7-+185.47 Lim (2.0-mg QIM;
`P<0.0001) for each group, respectively (Fig 4; Table 3).
`The reductions of leakage area, CNV area, and lesion size on
`FA compared with baseline were statistically significant (Fig 5).
`All
`types
`of
`neovascular AMD (classic,
`occult,
`and
`predominantly classic lesions) were included in the study, and
`after
`12 months of
`treatment,
`there were no
`significant
`differences between the 4 dosing groups with respect to changes
`in the lesions (P>0.05).
`Over the maintenance phase, the mean numbers of conbercept
`injections at 12 months were 4.73 (0.5-mg PRN group), 8.34 (0.5-
`mg Q1M group), 4.88 (2.0-mg PRN group), and 8.57 (2.0-mg
`QIM group). The 0.5-mg PRN group had 3.6 fewer injections
`than the 0.5-mg Q1M group, and the 2.0-mg PRN group had 3.7
`fewer injections than the 2.0-mg Q1M group. The study results
`confirmed that the PRN groups received significantly fewer in-
`jections than the QIM groups (P<0.05). The difference in the
`number of injections and the difference in the improvement of
`BCVAbetween both PRN groups were notstatistically significant
`(P>0.05).
`
`Safety
`
`tolerated. The most common
`Intravitreal conbercept was well
`ocular AEs that occurred were associated with intravitreal
`in-
`
`jections such as transient increased intraocular pressure, vitreous
`floaters,
`cataract,
`conjunctival
`hemorrhage,
`and
`corneal
`
`16 5
`14
`
`‘a
`
`13
`
`_12
`$10 |
`2
`|
`~ 3
`3
`2 6
`7
`4
`
`; ott
`
`0
`
`Oo
`oO
`0
`
`|_|
`s-15
`
`@0.5mg PRN
`0.5mg Q1M
`
`m 2.0mg PRN
`m2.0mg Q1M
`
`6
`
`3
`
`;
`
`5
`
`4
`
`3
`
`Hl
`
`E
`
`99
`
`88
`
`6
`
`4
`
`[
`
`215&<30
`20&<15
`>-15&<0
`Change of BCVA (Letters)
`
`230
`
`Figure 3. The number of patients with change in best-corrected visual
`acuity (BCVA) from baseline at month 12 in the 4 dosing regimens.
`PRN = prore nata (as needed); QIM = monthly.
`
`1744
`
`Regeneron Exhibit 2027
`Page 05 of 08
`
`Regeneron Exhibit 2027
`Page 05 of 08
`
`

`

`Li et al
`
`+ Conbercept for Neovascular AMD
`
`Table 3. Outcomes after the Maintenance Phase at Month 12
`
`Outcomes
`
`As Needed (n = 26)
`
`Monthly (n = 29)
`
`As Needed (n = 26)
`
`Monthly (n = 30)
`
`0.5-mg Group
`
`2.0-mg Group
`
`Increase of >30 letters, no. (%)
`Increase of 15—29 letters, no. (%)
`Increase of O—14 letters, no. (%)
`Decrease of 1—14 letters, no. (%)
`Decrease of >15 letters, no. (%)
`Change of CRT from BSL (ttm)
`Change in leakage area (mm)
`Change in CNV area (mm’)
`Changein lesionsize (mm?)
`No. of injections
`
`4 (15.4)
`9 (34.6)
`6 (23.1)
`7 (26.9)
`0 (0)
`—119.84+175.50
`—3.08+5.94
`—2.85+6.03
`—2.56+5.31
`7.73
`
`0 (0)
`9 (31.0)
`14 (48.3)
`5 (17.2)
`1 (3.4)
`—129.74170.80
`—4.76+5.90
`—3.8645.87
`—3.22+5.18
`11.34
`
`3 (11.5)
`8 (30.8)
`10 (38.5)
`4 (15.4)
`1 G.8)
`
`—152.14142.73
`—7.29+7.17
`—6.1145.99
`—5.7145.57
`7.88
`
`6 (20.0)
`8 (26.7)
`13 (43.3)
`3 (10.0)
`0 (0)
`
`—170.8+160.43
`—4.414+4.82
`
`—3.8845.21
`—4.12+5.77
`11.57
`
`BSL = baseline; CRT = central retinal thickness; CNV = choroidal neovascularization.
`
`Data are mean +standard deviation unless otherwise indicated.
`
`Discussion
`
`In the AURORA study, multiple intravitreal injections of
`conbercept resulted in a significant rapid increase of BCVA
`and a reduction of CRT, lesion area, and leakage in eyes of
`patients with neovascular AMD. Within the first 3 months,
`improvements in BCVA were evident after the 3 loading-
`dose injections in the 0.5- and 2.0-mg groups, and these
`improvements were maintained or
`increased through
`
`12 months by using a variable dosing regimen or fixed
`monthly dosing (0.5 mg PRN, 0.5 mg QIM, 2.0 mg PRN,
`and 2.0 mg QIM).
`The improvements in BCVA and anatomic outcomes in
`the 4 treatment groups were statistically significant when
`compared with baseline, but there were nostatistically sig-
`nificant differences between the 2 doses (0.5 and 2.0 mg) or
`the 2 dosing regimens (Q1M and PRN). In addition, there
`were no significant differences between dose and regimen
`
`
`
`
`
`7
`
`#
`
`#
`
`#
`
`oe
`
`tk
`
`Cc
`
`0 4
`
`1
`
`ee
`E
`~34
`—
`y
`Ww
`2-4 4
`2
`-5 4
`
`w 4
`
`64
`
`0
`
`-14
`‘2 4
`Ea
`o
`v-4
`a
`f° 4
`a-6 4
`3
`74
`3 J
`
`@0.5mg PRN
`w0.Smg QIM
`aZ,0ng EEN
`=2.0mg QIM
`
`*P <0.05
`"P< 0.01
`**P < 0,001
`
`0 -
`
`-14
`m0.5mg PRN
`m0.smgQIM 2-2 4
`2.0mg PRN
`E 3
`wa]
`_ m2.0mg QIM o a4
`= . |
`5°
`6 4
`7
`
`openue
`P< 0.01
`**D < 0,001
`
`Month
`
`:
`
`4
`
`+
`
`#
`
`#0.5mg PRN
`B0.5mg Q1IM
`m2.0mg PRN
`@2.0mg QIM
`
`en
`
`te
`
`*P <0.05
`
`4p <0,01
`**P <0.001
`
`Figure 5. The mean changes in (A) leakage area, (B) area of choroidal neovascularization (CNV), (C) and lesion size, determined using fluorescein
`angiography, from baseline in patients with 4 dosing regimens at months 3 and
`12. PRN = prore nata (as needed); QIM = monthly.
`
`1745
`
`Regeneron Exhibit 2027
`Page 06 of 08
`
`Regeneron Exhibit 2027
`Page 06 of 08
`
`

`

`Ophthalmology Volume 121, Number 9, September 2014
`
`Table 4. Serious Adverse Events and Most Frequent Treatment-
`Related Adverse Events during the Study
`
`
`
`2 (3.39)!
`(1.69)
`0
`
`(1.69)
`(1.69)
`0
`(1.69)
`(1.69)
`(1.69)
`(1.69)
`(1.69)
`
`7 (11.86)
`6 (10.17)
`(1.69)
`0
`(1.69)
`3 (5.08)
`(1.69)
`0
`0
`(1.69)
`0
`0
`
`0
`0
`1 (1.64)
`
`0
`0
`1 (1.64)
`0
`0
`0
`1 (1.64)
`0
`
`5 (8.20)
`3 (4.92)
`3 (4.92)
`4 (6.56)
`3 (4.92)
`0
`2 (3.28)
`2 (3.28)
`2 (3.28)
`0
`1 (1.64)
`1 (1.64)
`
`results from the AURORA study confirmed these earlier
`benefits from the phase 1 study and demonstrated that the
`benefits of intravitreal
`injections of conbercept could be
`SAEs/AEs 0.5-mg Group=2.0-mg Group
`maintained for at least 12 months.
`During the AURORAstudy,all digital images from OCT,
`FA, ICGA, and FP were sent to the independent masked
`central reading center known as the Digital Angiography
`Reading Center for analyses. Other studies, such as the
`EXCITE(the Efficacy and Safety of Ranibizumabin Patients
`With Subfoveal Choroidal Neovascularization [CNV]
`Secondary to Age-related Macular Degeneration [AMD])
`study,” which used ranibizumab, and the phase 1'° and
`phase 2 (CLinical Evaluation of Anti-angiogenesis in the
`Retina Intravitreal Trial [CLEAR-IT 1])'’ studies, which
`used aflibercept, used the Digital Angiography Reading
`Center as well. As a result, the image analyses performed
`in the AURORA(Assessthe safety and efficacy of KH902
`in patients with sUbfoveal choroidal neovasculaRization
`secOndary to age-Related mAcular degeneration) study
`were of the same standard as those in these otherstudies.
`Overall, conbercept waswell tolerated, and the incidence
`of ocular AEs was low. The most common AEs usually
`were caused by the intravitreal
`injection procedure and
`disappeared with or without treatment. The common intra-
`ocular and systemic AEs associated with conbercept also
`occurred with other similar anti-VEGF products,'* 7 such
`as the incidence of endophthalmitis,
`iritis, and decreased
`visual acuity. The SAEs, such as retinal detachment or
`laceration, that were reported with other similar products did
`not occur in this s