HIGH–MOLECULAR-WEIGHT
`AGGREGATES IN REPACKAGED
`BEVACIZUMAB
`
`MALIK Y. KAHOOK, MD,* LU LIU, MS,† PHILIP RUZYCKI, BS,*
`NARESH MANDAVA, MD,* JOHN F. CARPENTER, PHD,†
`J. MARK PETRASH, PHD,* DAVID A. AMMAR, PHD*
`
`Purpose: The antivascular endothelial growth factor agents ranibizumab and bevaci-
`zumab are used to treat ocular neovascular diseases. There have been recent reports of
`sustained elevation of intraocular pressure after use of either agent, which we hypothesize
`could be because of high–molecular-weight aggregates.
`Methods: Enzyme-linked immunosorbent assay, size exclusion chromatography, and
`polyacrylamide gel electrophoresis were used to analyze repackaged bevacizumab
`syringes obtained from three outside compounding pharmacies and samples obtained
`directly from the original vial. Microflow imaging was used to examine particulate material
`within samples.
`Results: All syringes contained statistically similar amounts of protein, consisting of
`immunoglobulin (IgG) heavy and light chains (polyacrylamide gel electrophoresis).
`However, two of the three compounding pharmacies’ batches had significantly less
`functional IgG in the solution (enzyme-linked immunosorbent assay). Additionally, the
`compounding pharmacies with the lowest IgG (;50%) also contained 10-fold the number
`of micron-sized particulate matter as measured by microflow imaging.
`Conclusion: There are significant differences in IgG concentration measured from
`repackaged bevacizumab syringes. A trend exists for an increase in micron-sized protein
`aggregates with the decrease in IgG concentration. Large particulate matter within some
`samples may lead to obstruction of aqueous outflow and subsequent elevation in
`intraocular pressure. Additional studies are warranted to explore these findings.
`RETINA 30:887–892, 2010
`
`T he most commonly used antivascular endothelial
`
`growth factor agents for ophthalmic neovascular
`diseases are bevacizumab and ranibizumab. Ranibizu-
`mab (Lucentis, Genentech, Inc., South San Francisco,
`CA), a Fab fragment of a recombinant humanized
`immunoglobulin-1 (IgG1) kappa isotype murine
`monoclonal antibody, is Food and Drug Administra-
`tion-approved specifically for treating wet age-related
`
`From the *Department of Ophthalmology, Rocky Mountain
`Lions Eye Institute, University of Colorado Denver, Aurora,
`Colorado; and the †Department of Pharmaceutical Sciences, Center
`for Pharmaceutical Biotechnology, University of Colorado Denver,
`Aurora, Colorado.
`M.Y.K. and N.M. have both received research support from
`Genentech in the past. No research support was received for this
`study.
`requests: Malik Y. Kahook, MD, Department of
`Reprint
`Ophthalmology, School of Medicine, University of Colorado
`Denver, 1675 Aurora Court, Mail Stop F 731, Aurora, CO 80045;
`e mail: malik.kahook@gmail.com
`
`macular degeneration.1 Bevacizumab (Avastin, Ge-
`nentech, Inc.), a humanized monoclonal antibody, is
`Food and Drug Administration-approved for treating
`colorectal cancer and is commonly used as an off-label
`treatment for ocular neovascular disease.2 Although
`the use of intravitreal bevacizumab (IVB) for ocular
`disease did not go through the Food and Drug Admin-
`istration approval process, published reports suggest
`that it is safe and apparently effective at treating wet
`age-related macular degeneration.3,4
`Despite the apparent safety profile of both IVB
`and ranibizumab, there have been recent reports of
`elevated intraocular pressure (IOP) after
`single
`or multiple injections of either agent.5–8 The number
`of reported cases of IOP spikes has, overwhelmingly,
`involved IVB prepared in plastic syringes by one or
`more repackaging formularies with fewer cases involv-
`ing ranibizumab obtained directly from the manufac-
`turer.6–7 In most cases, the affected patients had no
`
`887
`
`In
`
`Novartis Exhibit 2322.001
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`888
`
`RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2010  VOLUME 30  NUMBER 6
`
`history of ocular hypertension or glaucoma and were
`noted to have stable IOP before intravitreal injections.
`Often medical therapy, laser trabeculoplasty, and/or
`filtration surgery was required to decrease IOP to safe
`levels. In this report, we seek to investigate the
`hypothesis that repackaged bevacizumab may contain
`aggregated proteins or other particulate contaminants
`that could potentially lead to alterations in aqueous
`humor outflow facility with subsequent spikes in IOP.
`
`Materials and Methods
`
`Bevacizumab
`
`Repackaged bevacizumab (either 1.25 mg/0.05 mL
`or 2.5 mg/0.10 mL) was ordered from 3 external
`compounding pharmacies (CPs) and arrived in plastic
`syringes. Bevacizumab in its original glass vial was
`ordered through the University of Colorado Hospital
`(UCH) outpatient pharmacy and was drawn into
`plastic syringes immediately before assays were
`performed in our laboratory. Comparisons were made
`between the groups (n = 5 each) of syringes.
`
`Protein Content
`
`Total protein content was determined by the bicin-
`choninic acid method (Micro BCA Protein Assay Kit,
`Pierce/Thermo Scientific, Rockford, IL). Bevacizu-
`mab preparations (25 mg/mL) were diluted 1:100
`(0.25 mg/mL) in phosphate-buffered saline (8 g/L
`sodium chloride, 2.16 g/L sodium phosphate dibasic
`heptahydrate, 0.2 g/L potassium phosphate monobasic,
`and pH 7.4). A standard curve (0.025–1.5 mg/ mL)
`was prepared from a serial dilution of purified human
`IgG (Pierce/Thermo Scientific). The final 562 nm
`absorbance was determined using a Synergy 4 Multi-
`Mode Microplate Reader (BioTek, Winooski, VT).
`Protein concentrations were determined by fitting
`absorbance data to a linear regression of the standard
`curve absorbances. In each syringe, bevacizumab
`protein content was determined from the average of
`triplicate assays.
`
`Immunoglobulin Content
`
`Total IgG concentration was determined using the
`Easy-Titer Human IgG (H+L) Assay Kit (Pierce/
`Thermo Scientific). Bevacizumab formulations (25
`mg/mL) were diluted l:105
`(250 ng/mL)
`in the
`supplied dilution buffer. A standard curve from 15.6
`ng/mL to 500 ng/mL was prepared from a serial
`dilution of purified human IgG in dilution buffer. The
`final 410 absorbance was determined using a Synergy
`4 Multi-Mode Microplate Reader. IgG concentrations
`
`were determined by fitting absorbance data to a semi-
`log regression of the standard curve absorbances. In
`each syringe, bevacizumab IgG content was de-
`termined from the average of duplicate assays.
`
`Protein Molecular Weight Determination by
`Polyacrylamide Gel Electrophoresis
`
`Bevacizumab samples were separated on NuPAGE
`Novex 10% Bis-Tris Gels (Invitrogen) using MOPS
`running buffer (Invitrogen). Three micrograms of each
`sample were prepared in NuPAGE LDS sample buffer
`(Invitrogen) with or without dithiothreitol as a re-
`ducing agent and heated to 70°C for 10 minutes before
`loading. Gels were run for ;1.5 hours at 150 W and
`stained with Coomassie Blue R-250 to visualize
`protein bands.
`
`Size Exclusion Chromatography
`
`Chromatography of bevacizumab samples was
`performed by fast protein liquid chromatography
`using a Superose 6 10/300 GL column (GE Healthcare
`Bio-Sciences, Piscataway, NJ) operated at a flow
`rate of 0.5 mL/minutes of phosphate-buffered saline.
`Protein elution was monitored with an inline ultravi-
`olet absorbance detector, and fractions were collected
`throughout. Protein standards included blue dextran
`(2,000 kD), thyroglobulin (669 kD), ferritin (450 kD),
`aldolase (158 kD), and ovalbumin (45 kD).
`
`Microflow Imaging
`
`Microflow imaging (MFI) (DPA 4100, Brightwell
`Technologies Inc., Ontario, Canada) determines par-
`ticle size and number by imaging a fluid as it passes
`through flow cells using a digital camera and software
`filters. Bevacizumab samples were diluted 10-fold with
`bevacizumab placebo (50 mmol/L sodium phosphate,
`pH 6.25, 60 mg/mL a,a-trehalose dihydrate, and 0.4
`mg/mL polysorbate 20) before injection into the MFI
`flow cell. For control/baseline particle measurements,
`50 mL of bevacizumab placebo formulation was drawn
`into a 1-mL tuberculin syringe, expelled into an
`Eppendorf
`tube, and diluted 10-fold with more
`bevacizumab placebo formulation before injection.
`
`Results
`
`The protein content in the 5 syringes obtained from
`the university pharmacy (UCH) and the 3 CPs (CP1,
`CP2, and CP3) had an average protein content of 29.5,
`29.2, 30.6, and 30.6 mg/mL, respectively. The data are
`shown in Table 1. By using analysis of variance for
`comparison,
`there was no statistically significant
`
`Copyright© by Ophthalmic Communications Society, Inc. Unauthorized reproduction of this article is prohibited.
`
`Novartis Exhibit 2322.002
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`AGGREGATES IN REPACKAGED BEVACIZUMAB  KAHOOK ET AL
`
`889
`
`"' Cl
`
`-"'
`0
`0
`0
`N
`
`J,
`
`-
`-
`
`UCH
`CP3
`
`er,
`
`"'
`.,,
`Cl
`'°
`'°
`J,
`
`0
`
`"'
`.,,
`Cl
`..,
`..,
`J,
`
`"'
`.,,
`"'
`Cl
`Cl
`..,
`-"'
`00
`~ IC)
`J,
`J,
`r------
`peak
`
`Table 1. Assay of Bevacizumab Content
`
`Source
`
`Protein (mg/mL)
`
`IgG (mg/mL)
`
`UCH
`CP1
`CP2
`CP3
`
`*P , 0.01.
`†P , 0.000001.
`
`29.5 6 0.8
`29.2 6 1.2
`30.6 6 0.2
`30.6 6 1.7
`
`23.8 6 1.4
`21.4 6 3.1
`18.5 6 3.1*
`12.1 6 1.2†
`
`difference in total protein content among any of the
`groups of syringes (P = 0.14). In contrast, statistically
`significant differences were noted in the concentration
`of IgG between the UCH-obtained syringes and both
`CP2 (P , 0.01) and CP3 (P , 0.000001) but not
`between UCH and CP1 (P = 0.15). The average
`amount of IgG contained in the 5 syringes from the
`various pharmacies was determined to be: 23.8 mg/mL
`(UCH), 21.4 mg/mL (CP1), 18.5 mg/mL (CP2), and
`12.1 mg/mL (CP3) (Table 1). Compared with the
`labeled concentration of 25 mg/mL, the concentration
`of bevacizumab varied from 213% to 1% in the
`syringes from the UCH pharmacy, from 223% to 7%
`in the syringes from CP1, from 247% to 214% in the
`syringes from CP2, and from 259% to 248% in the
`syringes from CP3 (n = 5 for each).
`We first sought to resolve differences in protein
`content and IgG content by determining whether
`protein degradation had occurred in these samples. As
`shown in Figure 1, all samples of bevacizumab from
`the 4 pharmacies resolved into the heavy and light IgG
`chains (50 and 25 kD) as expected for IgG samples
`treated with a reducing agent such as dithiothreitol. No
`other bands were detected, and the intensity of the
`bands was approximately equal from lane to lane.
`These results suggest that lower IgG levels measured
`in some samples did not arise from protein fragmen-
`tation during repackaging and storage.
`
`6
`
`8
`
`18
`
`20
`
`22
`
`16
`14
`12
`10
`Elution Volume (ml)
`Fig. 2. Size exclusion chromatography of bevacizumab samples. Fifty
`micrograms of bevacizumab from UCH and CP3 were loaded onto a
`Superose 6 chromatography column. Elution volumes for protein stan
`dards are indicated by arrows above the graph. Fractions corresponding to
`the major peak and leading shoulder are indicated in boxes. Because of
`interaction with the Superose chromatography matrix, IgG molecules
`such as bevacizumab elute at a position corresponding to ;50 kD.
`
`Second, we used size exclusion chromatography to
`determine if large protein aggregates may have formed
`in the samples from the different CPs. Figure 2 shows
`size exclusion chromatography elution profiles of
`bevacizumab obtained from UCH (which shows
`normal IgG content) compared with a sample obtained
`from CP3 (which showed only ;50% of the normal
`IgG content). All bevacizumab samples demonstrated
`a major peak eluting at a volume expected for IgG
`
`Native
`GP1
`GP2
`
`UGH
`
`GP3
`
`UGH
`
`Reduced
`GP1
`GP2
`
`GP3
`
`Fig. 1. Polyacrylamide gel
`electrophoresis
`of
`bev
`acizumab samples. Three mi
`crograms from two syringes
`from each source of bev
`acizumab was run under na
`tive and reducing conditions.
`A single band was seen in the
`native gel, representing the
`full sized IgG (;150 kD).
`On denaturing, only the
`heavy (;50 kD) and light
`(;25 kD) IgG chains could
`be detected.
`
`-
`
`-
`
`-
`
`191 kDa
`
`97
`
`64
`
`-s,
`
`-
`
`-
`
`39
`
`28
`
`-19
`
`In
`
`Novartis Exhibit 2322.003
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`890
`
`RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2010  VOLUME 30  NUMBER 6
`
`Peak
`
`Shoulder
`
`UCH
`
`CP3
`
`UCH
`
`CP3
`
`UCH
`
`CP3
`
`UCH
`
`CP3
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`191 kDa
`
`97
`
`64
`
`s1
`
`39
`
`28
`
`19
`
`14
`
`Fig. 3. Polyacrylamide gel
`electrophoresis of size exclu
`sion chromatography frac
`tions. Denaturing and nonde
`naturing polyacrylamide gel
`electrophoreses were per
`formed on peak (left) and
`shoulder (right) fractions from
`the Superose separated UCH
`and CP3 bevacizumab sam
`ples. Peak fractions appeared
`identical to the gels of starting
`material. Shoulder fractions
`showed a high molecular
`weight aggregate (arrow) that
`resolves into heavy (;50 kD)
`and light (;25 kD) IgG chains
`on addition of a reducing agent.
`
`Native
`
`Reduced
`
`Native
`
`Reduced
`
`when analyzed under these conditions. In addition,
`a relatively minor shoulder at the leading edge of the
`peak was observed with both samples (Figure 2).
`Analysis by sodium dodecyl sulfate polyacrylamide
`gel electrophoresis under reducing conditions con-
`firmed that the peak and shoulder fractions contained
`IgG as evidenced by the characteristic pattern of heavy
`and light chains (Figure 3). When samples were
`analyzed without pretreatment with a reducing agent,
`aggregates of apparently higher molecular weight
`were noted (arrow, Figure 3) in fractions from the
`‘‘shoulder’’ samples but not the major peak. Because
`these higher molecular weight forms do not exist in
`the presence of a reducing agent but are stable by
`electrophoresis, they most likely arise from intermo-
`lecular disulfide crosslinks between two or more IgG
`molecules.
`In an effort to determine whether differences exist in
`the number and size of particles in different syringe
`groups, MFI was performed on four samples—one
`directly from the vial obtained from the UCH phar-
`macy and one each from all three CPs. The number of
`0.7 mm to 400 mm-sized particles in the single UCH
`sample tested was determined to be 61,000/mL,
`whereas the number of particles in the bevacizumab
`placebo alone (no protein) was 37,000/mL. The total
`number of particles from CP1 and CP2 was 57,000 and
`73,000, respectively. The total number of particles
`in the CP3 sample was 510,000/mL. These findings
`
`show that CP3 had almost 10 times the number of
`large (.1 mm) particulate matter when compared with
`the bevacizumab sample taken directly from the vial.
`Some of the particulate matter in CP3 reached a
`diameter of ;19 mm.
`
`Discussion
`
`Reports of persistent IOP spikes after single or
`multiple injections of bevacizumab have increased
`since the introduction of antivascular endothelial
`growth factor agents for treating wet age-related
`macular degeneration.5–8 We focused attention on
`bevacizumab because this agent is involved in most
`reported cases of IOP complications. Our investigation
`into potential causes of these IOP spikes showed
`significant variability in the concentration of bevaci-
`zumab IgG in some samples, although the total protein
`concentration was at expected levels in all samples
`examined. This led us to investigate the possibility that
`differences in levels of protein aggregation exist
`between the various sources of bevacizumab. As
`shown by size exclusion chromatography and poly-
`acrylamide gel electrophoresis analysis, all samples
`examined contained a population of high–molecular-
`weight aggregates that likely represented dimers or
`trimers of bevacizumab IgG monomers. To quantify
`the number of particles between different syringes, the
`
`Novartis Exhibit 2322.004
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`AGGREGATES IN REPACKAGED BEVACIZUMAB  KAHOOK ET AL
`
`891
`
`more sensitive MFI method was performed. This
`approach showed a substantially higher number of
`.1-mm particles in the material sourced from a CP
`when compared with material not prepared by a CP.
`The CP3 sample, which contained the highest number
`of .1-mm particles, was the same material that had
`the lowest concentration of the bevacizumab IgG.
`Of all the analytical studies shown here, MFI is the
`only 1 sensitive enough to detect these larger particles
`because: 1) although micron-sized particles contain
`millions of proteins, they represent a trace amount of
`the total protein and would be nearly invisible by poly-
`acrylamide gel electrophoresis; 2) the size of these
`particles would preclude them from entering the size
`exclusion chromatography column; and 3) in addition
`to protein, some of the particles could be resulting
`from nonprotein materials such as silicone oil from the
`syringes. These findings require further investigation.
`The introduction of ranibizumab-antibody-based
`therapeutics for ophthalmic disease is a relatively new
`phenomenon.10
`It
`spawned the off-label use of
`bevacizumab and created for the first time a need
`for CP-based repackaging of a recombinant protein
`therapy. With this new therapeutic group comes the
`need for a better understanding of how these agents
`should be handled by the CP and physicians alike. The
`importance of precise and reproducible handling of
`repackaged bevacizumab was
`introduced in the
`original Bascom Palmer protocol for using bevacizu-
`mab for intraocular injections, which covered storage
`conditions and expiration date based on work pub-
`lished by Bakri et al.11 Bakri et al found that the
`stability of bevacizumab repackaged into syringes
`documented a substantial loss of activity after ,3
`months (;8.8% loss) of refrigerated storage. Often
`in storage studies of therapeutic proteins, aggregate
`formation accounts for such loss of activity.12,13
`Conversion of ;10% of the monomeric protein into
`aggregates and particles could result in injection of
`a relatively high level of nonfunctional protein into the
`eye. Therefore, storage stability assessment for repack-
`aged bevacizumab should be based on the degradation
`profile as well as retained biologic activity.
`Aggregation of therapeutic proteins is ubiquitous
`during all stages of production from initial fermen-
`tation to purification and to filling of syringes and
`vials.12,13 Furthermore, aggregation occurs readily
`during shipping, storage, and even delivery of the drug
`product.12,13 Aggregation can be stimulated by con-
`ditions such as warming above refrigerator temper-
`atures, freeze–thawing, and excursions in pH.12,13 An
`extremely potent cause of aggregation is exposure to
`interfaces such as the air–water,
`liquid–solid, and
`liquid–silicone oil interfaces.14–16
`
`Furthermore, protein molecules readily adhere to
`microparticles of foreign materials (such as glass from
`vials or syringes, rubber from stoppers, stainless steel
`from filling pumps, and silicone oil used to lubricate
`syringes) creating protein particles and potentially
`seeding further aggregates.16,17
`Of particular importance for the current example of
`bevacizumab repackaging, a formulation that was
`originally developed to maintain protein stability in its
`original glass vial, is whether a plastic syringe is an
`appropriate container for preventing protein aggrega-
`tion and particle formation. In fact, there are examples
`in which a formulation of a protein therapeutic, Food
`and Drug Administration-approved for storage in
`a glass vial, failed to prevent protein aggregation when
`the same formulation was used in a prefilled syringe.16
`One of the main reasons for formulation failure in
`these conditions is that,
`in a syringe,
`the protein
`solution is exposed to materials and surfaces such as
`silicone oil that it may not encounter in a vial.16,17
`Another important issue is that storage of a protein
`solution in a plastic syringe can extract leachable
`material
`from the syringe barrel and the rubber
`plunger tip. A range of materials can enter the protein
`solution, including ions such as zinc, plasticizers, and
`other organic molecules.18,19 These compounds can
`induce protein aggregation as well as potentiate
`adverse effects in patients such as immunogenicity.19
`Therefore, properly developed therapeutic protein
`products are not stored in plastic containers.18
`In conclusion, our findings indicate that there are
`significant differences in IgG concentration between
`groups of IVB syringes. The reasons for these differ-
`ences require further studies, but a trend exists for an
`increase in micron-sized protein aggregates with the
`decrease in IgG concentration. Although the clinical
`use of IVB has demonstrated a broad therapeutic index
`of safety, it is unclear how the variability in expected
`concentration/dose that we identified may affect
`clinical outcomes. Lower than expected concentra-
`tions may diminish the robustness of efficacy or,
`alternatively, higher concentrations may lead to toxic
`effects or higher rates of uveitis.20 Obstruction of the
`outflow pathway by particulate matter with subsequent
`elevation in IOP is a well-known phenomenon both in
`disease states and in experimental glaucoma mod-
`els.21,22 The presence of protein aggregates and a high
`number of
`large molecules in some syringes of
`bevacizumab obtained from CPs, not seen in samples
`taken directly from original vials, could potentially
`lead to obstruction of the outflow pathways. We cannot
`at this time directly correlate protein aggregates with
`subsequent elevation in IOP that is resistant to medical
`therapy, and we cannot hypothesize whether single or
`
`Copyright© by Ophthalmic Communications Society, Inc. Unauthorized reproduction of this article is prohibited.
`
`Novartis Exhibit 2322.005
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`892
`
`RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2010  VOLUME 30  NUMBER 6
`
`multiple injections of aggregated protein are necessary
`to trigger IOP elevations. The lack of large-scale
`reports of IOP spikes post-IVB might indicate that
`alterations in the repackaged bevacizumab, similar to
`what we report, are only occurring in a few CPs with
`particular compounding habits. Therefore, physicians
`are encouraged to contact their respective CP and
`inquire about compounding techniques, duration of
`storage for IVB syringes, and what quality control
`measures are in place to assure that shipped syringes
`will arrive free of contaminants or alteration in the
`concentration of the active ingredient. Analysis of
`a larger number of syringes obtained from a greater
`number of CPs may help quantify the scope of this
`potential problem and help address issues that lead to
`variability. Additional studies are needed to explore
`the potential
`link between our findings and the
`observed clinical phenomenon of sustained elevation
`in IOP postinjection of repackaged bevacizumab.
`
`Key words: bevacizumab, ranibizumab, vascular
`endothelial growth factor, glaucoma,
`intraocular
`pressure, neovascular age-related macular degeneration,
`size exclusion chromatography, microflow imaging.
`
`References
`
`1. Bressler NM. Antiangiogenic approaches to age related mac
`ular degeneration today. Ophthalmology 2009;116:S15 S23.
`2. Ciulla TA, Rosenfeld PJ. Antivascular endothelial growth
`factor therapy for neovascular age related macular degenera
`tion. Curr Opin Ophthalmol 2009;20:158 165.
`3. Fung AE, Rosenfeld PJ, Reichel E. The International Intra
`vitreal Bevacizumab Safety Survey: using the internet to assess
`drug safety worldwide. Br J Ophthalmol 2006;90:1344 1349.
`4. Landa G, Amde W, Doshi V, et al. Comparative study of
`intravitreal bevacizumab (Avastin) versus ranibizumab (Lu
`centis) in the treatment of neovascular age related macular
`degeneration. Ophthalmologica 2009;223:370 375.
`5. Bakri SJ, McCannel CA, Edwards AO, Moshfeghi DM.
`Persistent ocular hypertension following intravitreal ranibizu
`mab. Graefes Arch Clin Exp Ophthalmol 2008;246:955 958.
`6. Kahook MY, Kimura AE, Wong LJ, Ammar DA, Maycotte MA,
`Mandava N. Sustained elevation in intraocular pressure
`associated with intravitreal bevacizumab injections. Ophthalmic
`Surg Lasers Imaging 2009;40:293 295.
`7. Carver J, Bouska C, Corey R. Avastin and Risk of Glaucoma.
`Retina Congress 2009 New York, September 30 October 4,
`2009. New York, NY. Abstract.
`
`Intravitreal bevacizumab
`8. Jalil A, Fenerty C, Charles S.
`(Avastin) causing acute glaucoma: an unreported complication.
`Eye 2007;21:1541.
`9. Michaelsen TE, Lofsgaard MF, Aase A, Heyman B. Un
`expected interaction of some anti TNP hybridoma antibodies
`with Superose HPLC gel filtration resins. J Immunol Methods
`1992;146:9 16.
`10. Heier JS, Antoszyk AN, Pavan PR, et al. Ranibizumab for
`treatment of neovascular age related macular degeneration:
`a phase I/II multicenter, controlled, multidose study. Ophthal
`mology 2006;113:633.el e4.
`11. Bakri SJ, Snyder MR, Pulido JS, McCannel CA, Weiss WT,
`Singh RJ. Six month stability of bevacizumab (Avastin)
`binding to vascular endothelial growth factor after withdrawal
`into a syringe and refrigeration or freezing. Retina 2006;26:
`519 522.
`12. Chi EY, Krishnan S, Randolph TW, Carpenter JF. Physical
`stability of proteins in aqueous solution: mechanism and
`driving forces in nonnative protein aggregation. Pharm Res
`2003;20:1325 1336.
`13. Mahler HC, Friess W, Grauschopf U, Kiese S. Protein
`aggregation: pathways, induction factors and analysis. J Pharm
`Sci 2009;98:2909 2934.
`14. Kerwin BA. Polysorbates 20 and 80 used in the formulation of
`protein biotherapeutics: structure and degradation pathways.
`J Pharm Sci 2008;97:2924 2935.
`15. Mahler HC, Senner F, Maeder K, Mueller R. Surface
`activity of a monoclonal antibody. J Pharm Sci 2009;98:
`4525 4533.
`16. Thirumangalathu R, Krishnan S, Ricci MS, Brems DN,
`Randolph TW, Carpenter JF. Silicone oil
`and agitation
`induced aggregation of a monoclonal antibody in aqueous
`solution. J Pharm Sci 2009;98:3167 3181.
`17. Bee JS, Chui D, Sawicki S, et al. Monoclonal antibody inter
`actions with micro and nanoparticles: adsorption, aggregation
`and accelerated stress studies. J Pharm Sci 2009;98:3218
`3238.
`18. Akers MJ, Nail SL. Development and Manufacture of Protein
`Pharmaceuticals. 1st ed. New York: Kluwer Academic/Plenum
`Publishers; 2002.
`19. Mueller R, Karle A, Vogt A, et al. Evaluation of the immuno
`stimulatory potential of stopper extractables and leachables by
`using dendritic cells as readout. J Pharm Sci 2009;98:3548
`3561.
`20. Georgopoulos M, Polak K, Prager F, Prunte C, Schmidt Erfurth U.
`Characteristics of severe intraocular inflammation following
`intravitreal injection of bevacizumab (Avastin). Br J Ophthalmol
`2009;93:457 462.
`21. Schwartz A. Chronic open angle glaucoma secondary to rheg
`matogenous retinal detachment. Trans Am Ophthalmol Soc
`1972;70:178 189.
`22. Weber AJ, Zelenak D. Experimental glaucoma in the primate
`induced by latex microspheres. J Neurosci Methods 2001;111:
`39 48.
`
`Copyright© by Ophthalmic Communications Society, Inc. Unauthorized reproduction of this article is prohibited.
`
`Novartis Exhibit 2322.006
`Regeneron v. Novartis, IPR2021-00816
`
`