Prospective Study of Intravitreal Ranibizumab as a
`Treatment for Decreased Visual Acuity Secondary to
`Central Retinal Vein Occlusion
`
`RICHARD F. SPAIDE, LOUIS K. CHANG, JAMES M. KLANCNIK, LAWRENCE A. YANNUZZI,
`JOHN SORENSON, JASON S. SLAKTER, K. BAILEY FREUND, AND ROBERT KLEIN
`
`● PURPOSE: To evaluate intravitreal injection of ranibi-
`zumab as a potential treatment for decreased visual acuity
`(VA) secondary to central retinal vein occlusion (CRVO).
`● DESIGN: Prospective, interventional case series.
`● METHODS: Patients with CRVO prospectively recruited
`from a practice were administered intravitreal ranibizumab 0.5
`mg (Lucentis; Genentech Inc, South San Francisco, Califor-
`nia, USA) at baseline and monthly for two additional doses.
`The patients were given additional ranibizumab if they had
`macular edema as determined by optical coherence tomogra-
`phy or any new intraretinal hemorrhage. Patients were evalu-
`ated for number of required injections, side effects, changes in
`VA, and macular thickness.
`● RESULTS: There were 20 eyes of 20 patients who at
`baseline had a mean age of 72.1 years, a mean VA of 45.8
`Early Treatment of Diabetic Retinopathy letters, and a
`mean central macular thickness of 574.6 ␮m. Of the 20
`eyes, five previously had received intravitreal triamcinolone
`and 11 had received intravitreal bevacizumab (Avastin;
`Genentech Inc). At 12 months of follow-up, the mean VA
`improved to 64.3 letters and the central macular thickness
`decreased to 186 ␮m (both different than baseline values; P <
`.001) using a mean of 8.5 injections. The change in macular
`thickness was not correlated with the change in VA. In one
`patient with a history of transient ischemic attack, an
`ischemic stroke developed but no sequela resulted. In
`another patient, vitreomacular traction developed, but the
`patient had improved acuity as compared with baseline.
`There were no infections, retinal tears, or detachments.
`● CONCLUSIONS: Intravitreal ranibizumab used over a period
`of one year improved mean VA, with low rates of adverse
`events, in patients with CRVO.
`(Am J Ophthalmol 2009;
`147:298–306. © 2009 by Elsevier Inc. All rights reserved.)
`
`C ENTRAL RETINAL VEIN OCCLUSION (CRVO) IS A
`
`condition characterized by dilation and tortuosity
`of the retinal venules with hemorrhages in all four
`quadrants of the retina. Often there is concomitant optic
`
`Accepted for publication Aug 10, 2008.
`From the Vitreous-Retina-Macula Consultants of New York and the
`LuEsther T. Mertz Retinal Research Center, Manhattan Eye, Ear, and
`Throat Hospital, New York, New York.
`Inquiries to Richard F. Spaide, Vitreous, Retina, Macula Consultants
`of New York, 460 Park Avenue, 5th Floor, New York, NY 10022; e-mail:
`rickspaide@yahoo.com
`
`nerve swelling and macular edema, and there may be
`exudative detachment of the retina, neovascularization of
`the iris, and neovascular glaucoma. After diabetic retinop-
`athy, CRVO is the second most common retinal vascular
`disease affecting the retina.1,2 Patients with CRVO report
`difficulty with many aspects of daily life and have decreased
`vision-related quality of life as measured by the 25-item
`National Eye Institute Visual Function Questionnaire.3
`Intravitreal levels of vascular endothelial growth factor
`(VEGF) in CRVO are the highest of those measured in
`retinal vascular disease,4 and the severity of findings in
`CRVO is proportional to intravitreal VEGF levels.5 Retinal
`neovascularization is seen rarely in eyes with CRVO. Almost
`all of the features of CRVO can be induced in primate eyes by
`intravitreal injection of VEGF, including vascular dilation,
`tortuosity, intraretinal hemorrhage, and capillary nonperfu-
`sion.6 Intravitreal VEGF injections in primates cause little or
`no retinal neovascularization.6 Inhibiting VEGF would seem
`to be a rational strategy for treating CRVO. Intravitreal
`injection of anti-VEGF agents caused regression of iris neo-
`vascularization.7,8 CRVO was among the first diseases treated
`with intravitreal bevacizumab (Avastin; Genentech Inc,
`South San Francisco, California, USA),9 and subsequent case
`series showed a large proportion of patients having visual
`acuity (VA) improvement.10,11 Bevacizumab treatment early
`after the onset of CRVO was associated with a statistically
`significant reduction in venous dilation, tortuosity, optic disc
`swelling, and macular edema in addition to improvement of
`VA.12 Ranibizumab (Lucentis; Genentech Inc) has poten-
`tially increased retinal penetration because of a smaller
`molecular size and a higher binding affinity for VEGF than
`does bevacizumab.13 Given these potential attributes, ranibi-
`zumab may be particularly suited as an agent to treat CRVO,
`a disease with manifestations that seem to be driven by
`VEGF. To gain preliminary information, we performed a
`prospective study of intravitreal ranibizumab injections as a
`treatment for decreased VA secondary to CRVO.
`
`METHODS
`
`THIS WAS A PROSPECTIVE, PHASE 1, OPEN-LABEL STUDY
`examining intravitreal injection of ranibizumab 0.5 mg as
`a treatment for decreased VA resulting from CRVO. The
`
`298
`
`© 2009 BY ELSEVIER INC. ALL RIGHTS RESERVED.
`
`0002-9394/09/$36.00
`doi:10.1016/j.ajo.2008.08.016
`
`Mylan Exhibit 1027
`Mylan v. Regeneron, IPR2021-00881
`Page 1
`
`Joining Petitioner: Apotex
`
`

`

`Ranibizumab for CRVO
`
`800
`
`700
`
`600
`
`100
`
`0 - (cid:173)
`
`-+Macubr Thickness
`.OORSLetters
`
`sasellne
`S75
`48.5
`
`90
`
`80
`
`'0
`
`20
`
`10
`
`6Month.s
`2S8
`S7.5
`
`58.4
`
`9 Months
`281
`S9.8
`
`12 Months
`!86
`61.l
`
`_ _,__
`FIGURE 1. Graph showing the thickness of the central macula
`in micrometers and the visual acuity (VA) in Early Treatment
`of Diabetic Retinopathy Study letters. CRVO = central retinal
`vein occlusion.
`
`study had Food and Drug Administration (FDA) Investi(cid:173)
`gator lnvestigational New Drug number 100,240.
`The patients with CRVO (as manifested by retinal
`hemorrhages and dilated retinal venules in all four quad(cid:173)
`rants of the fundus) underwent a comprehensive ophthal(cid:173)
`mologic examination that included VA assessment using
`Early Treatment of Diabetic Retinopathy (ETDRS) pro(cid:173)
`tocol refraction measured at 4 m, slit-lamp biomicroscopy,
`indirect ophthalmoscopy, color fundus photography, opti(cid:173)
`cal coherence tomography (OCT) using the Stratus OCT
`(Carl Zeiss Meditec, Dublin, California, USA), and fluo(cid:173)
`rescein angiography (FA). The central macular thickness
`(CMT) was measured manually using the software calipers
`because of the known inaccuracies of the segmentation
`algorithm with the Stratus OCT software. 14 Digital FA was
`performed with a T opcon T RC-50DX fundus camera with
`lmageNet software (Topcon USA; Paramus, New Jersey,
`USA). FA was performed by posit ioning the patient at the
`fundus camera, injecting 2 ml of a 25% solution of sodium
`fluorescein (Akom Inc, Somerset, New Jersey, USA) as a
`bolus into the antecubital vein. The timer on the camera
`was started at the time of injection, and filling, mid-phase,
`and late-phase images were recorded. The angiograms were
`evaluated for what was termed arm-to-early arterial filling,
`which was the time in seconds from the injection to when
`the dye front with in the superior and inferior temporal
`arcade arterioles both were seen to extend temporal to the
`retinal midline as defined by a vertical line extending
`through the geometric center of the fovea. Late-phase
`angiograms were evaluated for the presence of leakage and
`for the accumulation of dye within the retina.
`Patients with decreased VA secondary to CRVO were
`eligible for the study if they were 18 years or older and had a
`VA score of 34 letters (20/200) or more and 73 letters (20/40)
`or less as determined by the ETDRS VA protocol. To be
`eligible for the study, the patients must have had macular
`edema with a thickness of more than 250 µ m. Patients who
`were pregnant, had uncontrolled hypertension or diabetes,
`had greater than moderate nonproliferative diabetic retinop(cid:173)
`athy, uncontrolled glaucoma, OCT evidence of a vitreoreti(cid:173)
`nal interface abnormality, or any other additional ocular
`diseases that could compromise irreversibly the VA of the
`study eye were excluded from the study. Additionally, pa(cid:173)
`tients were not eligible for the study if they had a history of
`vitreous surgery or intravitreal, peribulbar, or retrobulbar
`corticosteroids within six months, cataract surgery within six
`months, yytrium- aluminum- garnet capsulotomy within two
`months, intravitreal injection of pegaptanib sodium within
`45 days, or intravitreal injection of bevacizumab within 30
`days of enrollment in the study.
`Enrolled patients were injected at baseline with 0.5 mg
`ranibizumab into the vitreous cavity on a monthly basis
`for the following two months and then were examined
`monthly. Patients were injected using aseptic technique as
`per the package labeling for ranibizumab, with the exception
`that no draping was used. Patients were examined within a
`
`period from three to seven days after injection at the
`beginning of the study, but this early examination was
`eliminated from the protocol after obtaining an exemption
`from the FDA approximately halfway through the study.
`Patients were examined on a monthly basis over the course of
`a one-year follow-up with ETDRS VA, slit-lamp biomicros(cid:173)
`copy, indirect ophthalmoscopy, color fundus photography,
`OCT, and FA Patients were reinjected with 0.5 mg ranibi(cid:173)
`zumab if they had evidence of any hemorrhage in the retina,
`or macular edema as determined if there was the presence of
`a macular thickness of more than 250 µm or intraretinal
`cystoid spaces within the macular region.
`
`• ST A TISTICAL METHODS: The data obtained were ana(cid:173)
`lyzed with frequency and descriptive statistics. Categorical
`analysis was performed using the Chi-square test. The
`Pearson correlat ion coefficient was calculated as a measure
`of the bivariate linear association between variables. Base(cid:173)
`line VA as measured in ETDRS letters was compared with
`the measurements obtained at th ree and six mon ths with a
`paired-samples t test. A P value of less than .05 was
`considered significant, except for the VA data, because two
`evaluations, one at three months and another at six
`months, were made. (The a was adjusted as per the
`Bonferroni technique.) There was case-wise deletion of miss(cid:173)
`ing data. Statistical analysis was performed with SPSS soft(cid:173)
`ware version 16.01 (SPSS Inc, Chicago, Illinois, USA).
`
`• EARLY EXIT FROM STUDY: Four patients exited the study
`before the 12-month follow-up. One patient found the travel
`necessary for the study visits too difficult, another patient
`moved to a different state for employment reasons and could
`not return for examinations. In one patient, focal vitreomacu(cid:173)
`lar traction developed and the patient underwent vitrectomy,
`and another patient was presumed to have a stroke, after
`which she did not want to receive any additional injections.
`The latter two patients are detailed in the Results. Exited
`
`VOL. 147, No. 2
`
`RANIBIZUMAB AND CRVO
`
`299
`
`Mylan Exhibit 1027
`Mylan v. Regeneron, IPR2021-00881
`Page 2
`
`Joining Petitioner: Apotex
`
`

`

`FIGURE 2. Images obtained from a 53-year-old male with a one-day history of CRVO. (Top left) Fundus photograph showing optic
`nerve head hyperemia and swelling, dilated tortuous veins, intraretinal hemorrhage, and foveal pseudocyst. (Top right) Fundus
`photograph obtained at the one-year follow-up showing the manifestations of the CRVO resolved. (Middle left and right) Optical
`coherence tomography (OCT) images obtained at baseline and at the one-year of follow-up showing dramatic resolution of the
`macular edema. (Bottom left) Midphase fluorescein angiogram (FA) showing a normal appearance grossly. (Bottom right) However
`on closer examination, there was a moth-eaten appearance of the capillary bed with very small areas of nonperfusion (arrows). The
`patient’s VA improved 22 letters.
`
`patients were not included in any calculations past the
`examinations they completed.
`
`RESULTS
`
`● BASELINE CHARACTERISTICS: There were 20 eyes of 20
`patients who at baseline had a median age of 72.1 years
`(interquartile range, 60.2 to 80.1 years). Twelve of the
`patients were male. Of the 20 patients 14 (70%) had
`hypertension, two (10%) had diabetes mellitus, two (10%)
`had a history of myocardial infarction, one (5%) had a prior
`history of a transient ischemic attack, one (5%) was a current
`smoker, and seven (35%) were former smokers. The median
`duration of the CRVO on entry to the study was 34.6 weeks
`(interquartile range, six to 75.5 weeks). Of the 20 patients, 11
`
`had previous treatment, five had previous intravitreal triam-
`cinolone treatment, and 11 (including all of those who had
`intravitreal triamcinolone treatment) had intravitreal bevaci-
`zumab treatment. The median number of past injections for
`those who received intravitreal bevacizumab was six (inter-
`quartile range, 4 to 9; mean, 6.2), with the dose being either
`1.25 or 2.5 mg. The median time from the last bevacizumab
`injection until entry into the study was 58 days (interquartile
`range, 36.8 to 89 days). The mean baseline VA was 45.8
`ETDRS letters (⫾ 17.7 [standard deviation (SD)]; median, 54
`letters) and the mean CMT was 574.6 ␮m (⫾ 181 [SD];
`median, 581 ␮m). The baseline VA and CMT did not differ
`by previous history of bevacizumab injection (P ⫽ .47 and
`P ⫽ .78, respectively). The FAs showed an arm-to-early
`arterial filling time occurring at a mean of 28.1 seconds after
`injection (range, 18.7 to 51.9 seconds). Retinochoroidal
`
`300
`
`AMERICAN JOURNAL OF OPHTHALMOLOGY
`
`FEBRUARY 2009
`
`Mylan Exhibit 1027
`Mylan v. Regeneron, IPR2021-00881
`Page 3
`
`Joining Petitioner: Apotex
`
`

`

`FIGURE 3. Images obtained from a 64-year-old man with a 52-week history of CRVO for which he had not received treatment.
`(Top left) FA obtained 32.5 seconds after an antecubital injection of fluorescein showing that the dye front is first becoming visible
`in the eye. (Top right) FA obtained at 33.9 seconds after dye injection showing that the dye front in both the inferior and superior
`arcade vessels has moved temporal to the midline through the center of the fovea. (Middle left) OCT image showing that the patient
`had macular edema at baseline. (Middle right) OCT image showing that at the 12-month follow-up, there was no macular edema.
`(Bottom left) FA obtained at the 12-month follow-up showing the dye front first becoming visible on the vessels on the nerve head
`at 32.3 seconds. (Bottom right) FA obtained at the 12-month follow-up showing that the dye front moved past the midline at 34.8
`seconds after injection. The patient’s VA improved 29 letters.
`
`collateral vessels were seen in 11 eyes and were not seen in
`three eyes, and determination was not possible in six eyes
`because of hemorrhage on or around the nerve.
`
`● EARLY FOLLOW-UP: In the follow-up during the first
`week after the first injection, the mean VA improved 8.3
`letters (P ⬍ .001) and the CMT decreased by a mean of
`349.2 to 225.4 ␮m of thickness (P ⬍ .001). The baseline
`fluorescein arm-to-early arterial filling time was correlated
`to change in VA at one week (r ⫽ 0.59; P ⫽ .015), but did
`not show a significant correlation with the change in CMT
`(r ⫽ 0.27; P ⫽ .3) at one week. At one month, the vision
`and thickness parameters regressed to a VA improvement
`
`of 6.2 letters over baseline and a CMT of 273 ␮m (P ⫽
`.019 and P ⬍ .001, respectively, as compared with base-
`line). One patient had a four-letter gain at one week, but
`an 18-letter loss at one month. He ultimately received
`nine injections over 12 months and had a 15-letter gain.
`At the three-month follow-up, the VA showed a mean
`improvement of 10.4 letters over baseline (P ⫽ .001), and
`there was a decrease of the CMT to 199 ␮m (P ⬍ .001).
`
`● TWELVE-MONTH FOLLOW-UP: At 12 months of follow-
`up, the mean VA improved to 64.3 letters (median, 68.5
`letters), and the CMT decreased to 186 ␮m (median, 170
`␮m; P ⬍ .001 for each; Figure 1). The patients received a
`
`VOL. 147, NO. 2
`
`RANIBIZUMAB AND CRVO
`
`301
`
`Mylan Exhibit 1027
`Mylan v. Regeneron, IPR2021-00881
`Page 4
`
`Joining Petitioner: Apotex
`
`

`

`FIGURE 4. (Top left) FA obtained at the 12-month follow-up examination showing that the manifestations of the CRVO have
`resolved except for a small number of scattered intraretinal hemorrhages (arrows) in this 73-year-old male. At 35 seconds after the
`antecubital injection of fluorescein, the dye front is visible in the arterioles of the vascular arcade (arrowheads). (Top right) FA
`showing that the dye front has moved distally within the vessels (arrowheads) by 38.6 seconds. (Bottom left) FA showing that the
`dye front has moved distally within the vessels (arrowheads) by 43.6 seconds after injection. (Bottom right) FA showing that the
`dye front has nearly moved out of the frame (arrowheads) by 47.9 seconds after injection. The patient’s VA improved six letters.
`
`mean number of 8.5 injections (median, 10). Considering
`the patients who exited early from the study, the injection
`frequency was 0.85 per month over the total potential
`months in which injection was possible, with 47.3% of the
`injections given for edema and 52.7% given principally
`because of hemorrhage criteria. The change in VA was not
`correlated with the number of injections (r ⫽ 0.39; P ⫽
`.14). Although both the VA and the CMT improved, the
`two measures were not correlated (r ⫽ 0.24; P ⫽ .38).
`There were nine (56.3%) eyes that had improvement of
`three lines or more, and two eyes (12.5%) that had an
`improvement of more than six lines of VA. The mean last
`observation carried forward acuity was 63 letters. There
`was one eye (6.3%) with a loss of three lines or more at 12
`months (that eye had a 20-letter loss). This eye had
`recurrence of macular edema after one month without an
`injection of ranibizumab when the acuity was 21 letters
`better. As part of an extension study, this patient had an
`injection of ranibizumab at 12 months and had a return of
`vision during that study (data not shown).
`The change in VA at one week was correlated with the
`change in acuity at 12 months (r ⫽ 0.56; P ⫽ .029), but the
`change at one month was not correlated with the change in
`acuity at 12 months (r ⫽ 0.37; P ⫽ .16). The change in CMT
`at one week was correlated with the change in macular
`
`thickness at 12 months (r ⫽ 0.94; P ⬍ .001). The change in
`CMT at one month showed a smaller correlation with
`the change in CMT at 12 months (r ⫽ 0.5; P ⫽ .047). The
`change in VA at 12 months was not correlated with the
`arm-to-early arterial filling time at baseline (r ⫽ 0.11; P ⫽
`.69). There was no difference in the VA change among
`those who had previous bevacizumab treatment vs those
`who did not (P ⫽ .61). In one patient, retinochoroidal
`collateral vessels developed, and in none of the patients
`with retinochoroidal collateral vessels at baseline did they
`disappear by 12 months. The mean arm-to-early arterial
`filling time at the final examination was 29.2 seconds, a
`change from baseline that was not significant (P ⫽ .9).
`Although not numerically quantified, cotton wool spots
`were seen to take many months to resolve in this series. At
`12 months, examination of the FAs showed capillary
`network in the retina had a moth-eaten appearance with
`telangiectasis (Figures 2 through 5).
`
`● SAFETY: There were a total of 170 intravitreal injec-
`tions over the study period, and there were no retinal tears,
`detachments, or infections resulting from these injections.
`One 84-year-old hypertensive woman had a cerebral vas-
`cular accident, or what was thought to be an ischemic
`cerebral vascular accident. She had a past history of
`
`302
`
`AMERICAN JOURNAL OF OPHTHALMOLOGY
`
`FEBRUARY 2009
`
`Mylan Exhibit 1027
`Mylan v. Regeneron, IPR2021-00881
`Page 5
`
`Joining Petitioner: Apotex
`
`

`

`FIGURE 5. Images obtained from a 74-year-old man with CRVO 35 weeks before entry to the study. In that period, he received
`five injections of bevacizumab with a dramatic improvement in retinal appearance. However he continued to have macular edema,
`dot hemorrhages in the retina, and moderate optic atrophy. (Top left) Fundus photograph showing the dot hemorrhages in the retina.
`(Top right) OCT images showing the macular edema. (Bottom left) Fundus photograph obtained after one year in the study showing
`that the patient had resolution of the venous dilation and intraretinal hemorrhage. (Bottom right) OCT images showing that the
`macular edema resolved. Note the retinal atrophy, most markedly seen involving the inner retina (arrow) of the parafoveal macula,
`which ordinarily is much thicker than the fovea itself (arrowhead). The patient’s VA improved 25 letters.
`
`transient unilateral weakness before entry in the study and
`was evaluated with computed tomography. No abnormal-
`ities were seen and the patient was thought to have had a
`transient ischemic attack. After the fourth month of
`follow-up while enrolled in this study, she had an episode
`of mild drooping of the corner of her mouth and weakness
`in her arm that lasted a few minutes. She underwent a
`computed tomography evaluation that found no abnor-
`mality, and she was discharged with the diagnosis of a
`transient ischemic attack. However, a magnetic reso-
`nance imaging scan performed as an outpatient found an
`infarct. She was discovered to have an ipsilateral redun-
`dant loop in her carotid artery, but did not have any
`alteration in arterial diameter. The patient had no observ-
`able sequela from her stroke.
`A 76-year-old male had a baseline VA of 53 letters and
`a baseline foveal thickness of 390 ␮m. Before entry into
`the study, he had undergone 10 injections of bevacizumab
`with partial resolution of his macular edema. He received
`two injections of ranibizumab in the study and had an
`increase in his VA to 60 letters and a decrease in macular
`thickness to 158 ␮m. However, he reported new-onset
`
`distortion and was seen to have typical findings of focal
`vitreomacular traction syndrome during OCT examina-
`tion. The patient underwent an uncomplicated vitrectomy
`with removal of the traction and removal of his internal
`limiting membrane. Because of the vitrectomy, he was
`exited from the study. Thereafter, he was treated using
`similar criteria to the study protocol, but with injections of
`bevacizumab. Curiously, over the course of the following
`year, the patient required only two bevacizumab injections.
`
`DISCUSSION
`
`THIS STUDY OF INTRAVITREAL RANIBIZUMAB FOR CRVO
`found an improvement in VA that began less than one
`week after the first injection and continued to show
`improvement after additional injections over the following
`year of follow-up. The patients showed a significant im-
`provement in VA, with 94% having stabilization or
`improvement in acuity at one year. There was a reduction
`in macular edema, but the amount of change in macular
`edema was not correlated with the amount of change in VA.
`
`VOL. 147, NO. 2
`
`RANIBIZUMAB AND CRVO
`
`303
`
`Mylan Exhibit 1027
`Mylan v. Regeneron, IPR2021-00881
`Page 6
`
`Joining Petitioner: Apotex
`
`

`

`Two patients had noteworthy side effects. In one patient
`with a previously unknown carotid artery abnormality but
`a past history of transient ischemic attack, what was
`thought to be another transient ischemic attack took place
`while the patient was participating in the study. A com-
`puted tomography scan obtained after each event showed
`no abnormality, but after the second transient ischemic
`attack, the patient underwent magnetic resonance imaging
`and magnetic resonance angiography, which indicated an
`infarct, probably of recent origin ipsilateral to a carotid
`artery abnormality. A potential concern with using any
`anti-VEGF agent is the theoretical increase in arterial
`thromboembolic events. A review of Medicare standard
`analytic files for patients with new-onset age-related mac-
`ular degeneration before the advent of anti-VEGF therapy
`found that patients with past history of arterial thrombo-
`embolic events had an annual prevalence rate of 7.4% for
`myocardial infarction and 35.1% for inpatient ischemic
`stroke,15 implying that these patients have a high under-
`lying risk for an arterial thromboembolic event. Although
`the rates for patients with age-related macular degenera-
`tion may be different than those with CRVO, a CRVO
`can be considered an ischemic event affecting the central
`nervous system, or at least a structure derived from the
`central nervous system. The second complication was the
`development of vitreomacular traction in a patient who
`had 10 previous bevacizumab injections and two injections
`of ranibizumab. Vitreomacular traction is a rare complication
`of CRVO16 and is associated more frequently with branch
`retinal vein occlusion.17–19 This patient underwent vitrec-
`tomy with internal limiting membrane removal and thereaf-
`ter seemed to require less frequent anti-VEGF injections,
`albeit with a different agent. Vitrectomy increases the oxygen
`concentration at the level of the retina,20 potentially reduc-
`ing the stimulus for VEGF production.
`In this series, blocking intraocular VEGF seemed to
`result in improved VA in patients with CRVO. Previous
`authors have classified CRVO as being ischemic or non-
`ischemic21 or, in the CRVO Study, as perfused vs nonper-
`fused (with indeterminate eyes not being immediately
`classifiable).22,23 The CRVO Study used FA evidence of
`areas of absent flow to define nonperfusion, principally to
`create a method to predict those patients in whom iris
`neovascularization would develop. Nonperfused eyes were
`found to have a greater risk for iris neovascularization.
`More recent studies using primate eyes found that intrav-
`itreal VEGF injections produce findings that emulate
`CRVO, including areas of nonperfusion.6,24 Histopatho-
`logic analysis of the injected eyes showed endothelial cell
`proliferation occurring in capillaries to the point where the
`entire capillary was occluded. Vascular endothelial prolif-
`eration occurred within venules leading to obstruction,
`and the regions drained by these venules were seen to have
`capillary nonperfusion during FA.24 It is likely that what
`appears as a nonperfused CRVO in humans is more related
`to the effects of VEGF and not something intrinsically
`
`specific to occlusion of the central retinal vein. It is
`possible that anti-VEGF agents may prevent or reverse the
`excessive intravascular proliferation of vascular endothe-
`lial cells. It is not logical to call an acute condition
`nonischemic when it was produced by a vascular occlusion
`that causes VEGF secretion from the affected tissue.
`CRVOs likely exist in a continuum from mild to severe,
`with the resultant secretion of VEGF from the retina
`increasing with worsening of the blood flow. At very high
`levels of VEGF, decompensation of the retinal vasculature
`from intravascular proliferation of vascular endothelial
`cells, as was seen in the primate experiments, creates a
`self-reinforcing cycle of decompensation.
`Intravitreal administration of anti-VEGF agents reduces
`tissue effects of VEGF, but does not necessarily address the
`underlying factors that lead to the increased VEGF in the
`first place. The arm-to-early arterial filling time was pro-
`longed in the patients at presentation (Figure 3). One
`possible explanation is that low flow and increased hydro-
`static pressure within the venules in turn caused a general
`slowing of the arterial inflow. However, after one year of
`receiving intravitreal ranibizumab with near normalization
`of the retinal thickness, retinal appearance, and improve-
`ment in VA, the arm-to-early arterial filling time showed
`no change (Figures 3 and 4). This finding implies that the
`arterial supply into these eyes was compromised. Reduc-
`tion of venous outflow with simple ligation of the vein in
`primates did not produce a picture consistent with severe
`CRVO; to produce severe CRVO, simultaneous compro-
`mise in arterial flow was required.25
`It is likely that poor arterial perfusion of the retina sets the
`stage for the manifestations we ascribe to a CRVO, which are
`mediated to a large part by VEGF. Inhibition of VEGF seems
`to reverse to pathologic decompensation within the retina,
`but is there a cost to the retina for this type of treatment?
`VEGF has been found to be an important survival factor for
`retinal neurons and is a critical neuroprotectant.26 Intuitively,
`it may seem that in this situation a nearly complete suppres-
`sion of VEGF-mediated effects by ranibizumab may be unde-
`sirable. However in our series, slightly more than half of the
`patients previously had undergone treatment with bevaci-
`zumab, an agent that may not block the effects of VEGF as
`well as ranibizumab, and these patients showed the same
`magnitude of improvement in vision as those not treated with
`prior bevacizumab (Figure 5). It is possible that withdrawal of
`VEGF and its cell survival and neuroprotective effects allow
`marginally oxygenated neurons actually to die, resulting in a
`state where the retina overall requires less oxygenation. The
`metabolic and oxygen demands of the remaining cells are
`more closely matched with what is available from the retinal
`circulation. However, the patients in the present study still
`required frequent injections, indicating there was an unre-
`solved perfusion deficit. This raises the possibility that treat-
`ing patients with ranibizumab may stabilize the retina to the
`extent that other procedures such as vitrectomy may have a
`favorable long-term effect.
`
`304
`
`AMERICAN JOURNAL OF OPHTHALMOLOGY
`
`FEBRUARY 2009
`
`Mylan Exhibit 1027
`Mylan v. Regeneron, IPR2021-00881
`Page 7
`
`Joining Petitioner: Apotex
`
`

`

`This series, although prospective, has a number of
`potential defects. The sample size was small, there was no
`control group, the investigators and patients were not
`masked, and several patients exited early from the study.
`Each of these factors limits the generalizability of the
`study. The use of ranibizumab in CRVO currently is being
`investigated in a randomized trial. The present study found
`that patients with CRVO showed a significant and sustain-
`able improvement in VA when treated with intravitreal
`ranibizumab following a set of criteria that called for
`treatment if any aspect of CRVO was present including
`edema, as well as the presence of intraretinal hemorrhage.
`
`This strategy was aimed at not solely treating macular
`edema. The strategy resulted in frequent injections over
`the year and resulted in a demonstrable improvement in
`vision. Of interest is that the VA at one month was not
`predictive of the one-year outcome, so a single injection
`followed by a wait-and-see approach is likely to be a
`suboptimal strategy. It may be possible that the best
`strategy is a recurrent periodic injection given at intervals
`found to prevent any recrudescence of retinal pathologic
`features, much the same as the strategy used in the
`treat-and-extend approach as has been described for cho-
`roidal neovascularization.27
`
`THIS STUDY WAS SUPPORTED BY A GRANT FROM GENENTECH INC, SOUTH SAN FRANCISCO, CALIFORNIA AND THE LUESTHER
`T. Mertz Retinal Research Center, Manhattan Eye, Ear, and Throat Hospital, New York, New York. Dr Spaide is a consultant to Topcon Inc and
`received research funding from Genentech Inc. Dr Chang is supported in part by the Heed Ophthalmic Foundation. Involved in design and conduct
`of study (R.F.S.); collection (R.F.S., L.K.C., J.M.K., J.S., J.S.S., K.B.F.), management (R.F.S., L.K.C.), analysis (R.F.S.), and interpretation of data
`(R.F.S.); and preparation (R.F.S.) and final approval of the manuscript (R.F.S., L.K.C., J.M.K., L.A.Y., J.S., J.S.S., K.B.F., R.K.). Institutional Review
`Board (IRB) approval through Western IRB. This study complied with the Health Insurance Portability and Accountability Act of 1996 and has the
`ClinicalTrials.gov registration number NCT00403039.
`
`REFERENCES
`
`1. Klein R, Moss SE, Meuer SM, Klein BE. The 15-year
`cumulative incidence of retinal vein occlusion: the Beaver
`Dam Eye Study. Arch Ophthalmol 2008;126:513–518.
`2. Cugati S, Wang JJ, Rochtchina E, Mitchell P. Ten-year inci-
`dence of retinal vein occlusion in an older population: the Blue
`Mountains Eye Study. Arch Ophthalmol. 2006;124:726–732.
`3. Deramo VA, Cox TA, Syed AB, Lee PP, Fekrat S. Vision-
`related quality of life in people with central retinal vein
`occlusion using the 25-item National Eye Institute Visual
`Function Questionnaire. Arch Ophthalmol 2003;121:1297–
`1302.
`4. Aiello LP, Avery RL, Arrigg PG, et al. Vascular endothelial
`growth factor in ocular fluid of patients with diabetic
`retinopathy and other retinal disorders. N Engl J Med
`1994;331:1480 –1487.
`5. Boyd SR, Zachary I, Chakravarthy U, et al. Correlation of
`increased vascular endothelial growth factor with neovascu-
`larization and permeability