AAPS PharmSciTech, Vol. 12, No. 2, .Tune 2011 (© 2011)
`DOI: 10.1208/sl2249-0ll-9617-y
`
`Research Article
`Theme: Sterile Products: Advances and Challenges in Formulation, Manufacturing,Devices and Regulatory Aspects
`Guest Editors: Lavinia Lewis, .Tim Agalloco, Bill Lambert, Russell Madsen, Mark Staples
`
`Development of Biotechnology Products in Pre-filled Syringes: Technical
`Considerations and Approaches
`
`Advait Badkar,1
`
`2 Amanda Wolf,1 Leigh Bohack,1 and Parag Kolhe1
`•
`
`Received 15 September 2010; accepted 14 April 2011; published online 4 May 2011
`
`Abstract. A monoclonal antibody (mAb) product development case study is presented to address some
`of the issues faced during developing a pre-filled syringe (PFS) product for a biotherapeutic. In
`particular, issues involving incompatibility with silicone oil and a stability-based approach for selection of
`PFS barrel and tip cap components have been discussed. Silicone spiking studies followed by exposure to
`agitation stress or accelerated temperature conditions were used to check for incompatibilities of the
`mAb with silicone oil, a necessary product contact material in PFS. In addition, screening studies to
`compare various closure materials as well as syringe barrel processing methods were used to select the
`optimum closure materials as well as the correct syringe processing method. Results indicate that the
`model mAb formulation used was sensitive to high levels of silicone oil especially under accelerated
`temperature conditions resulting in formation of protein-silicone particles in the solution for samples that
`were spiked with the silicone oil. Agitation stress did not have any significant impact on the quality
`attributes tested. Samples stored in syringe barrels that were processed with sprayed-on silicone had
`higher levels of subvisible particles as compared to those that were processed with the baked-on process.
`The tip cap comparability study resulted in one tip cap material having superior compatibility among the
`three that were tested. The quality attribute that was most impacted by the tip cap materials was mAb
`oxidation. An approach for evaluation of primary packaging components during the development of pre(cid:173)
`filled syringe presentations for biotechnology-based compounds has been highlighted.
`
`KEY WORDS: formulation development; PFS; pre-filled syringes; primary container closure system;
`protein development; silicone oil compatibility; syringe processing; tip cap selection.
`
`INTRODUCTION
`
`The development of biotechnology products such as
`therapeutic proteins or vaccines in pre-filled syringes (PFS)
`has attracted considerable interest from pharmaceutical and
`biotechnology companies in the past 5 to 7 years. This can be
`attributed mainly to several benefits that PFS provide
`including reduction of medical dosing errors, minimized risk
`for microbial contamination by limiting manipulations prior
`to dosing, and enhanced convenience and ease of use (1).
`Pre-filled syringes represent a win-win scenario for end-users
`as well as manufacturers as they encourage and simplify self(cid:173)
`administration by the patient or caregiver and provide
`reduced therapy and injection cost. A lower "dead volume"
`in PFS results in a greater number of filled units for a given
`batch compared to vials hence increasing overall process yield
`during manufacturing.
`
`1 Pharmaceutical Research & Development, Biotherapeutics Phar(cid:173)
`maceutical Sciences, Pfizer Inc, 700 Chesterfield Parkway West (AA
`3I), Chesterfield, Missouri 63017, USA.
`2 To whom correspondence should be addressed. ( e-mail: advait.v.
`badkar@pfizer.com)
`
`In spite of these aforementioned advantages, the devel(cid:173)
`opment of biotechnology products in PFS presents its own set
`of distinctive challenges. Most of them are due to the complex
`and amphiphilic nature of proteins and impact of the
`container closure on protein stability. Protein stability in
`PFS over shelf life is a concern due to potential incompati(cid:173)
`bility of these types of molecules with various contact surfaces
`and interactions with leachables of the PFS system. Prolonged
`exposure of these leachables can also increase the potential
`for various forms of protein degradation including conforma(cid:173)
`tional changes, reduced activity, and increased immunogenic(cid:173)
`ity. In addition to protein stability, there are several other
`technical challenges that are faced by manufacturers includ(cid:173)
`ing functionality issues (break loose, glide force, and tip cap
`removal force), filling issues such as "wet plungers", plunger
`movement during shipping, and container closure integrity of
`PFS system.
`In this paper, we have highlighted an approach that we
`used for assessing the suitability of a PFS system for
`development of a monoclonal antibody (mAb) product. In
`particular, issues involving incompatibility with silicone oil
`and a stability-based approach for selection of PFS barrel and
`tip cap components have been discussed.
`
`1530-9932/11/0200-0564/0 © 2011 American Association of Pharmaceutical Scientists
`
`564
`
`"' aaps·
`
`Regeneron Exhibit 1044.001
`
`

`

`Development of Biotechnology Products in Pre-filled Syringes
`
`565
`
`PFS systems essentially consist of a glass barrel made up
`of type I borosilicate glass that is typically siliconized for
`syringe functionality (2). Several vendors have the option of a
`non-glass (plastic) barrel available, but this technology has
`not attracted a lot of interest and use of glass barrels is still
`dominant in the industry. Typically during the barrel for(cid:173)
`mation, the glass cane is cut into the required length and
`formed into the flange end and tip (luer tip). A tungsten pin
`( other newer metal pins are being explored as well) is used to
`bore the fluid path through the glass barrel tip. In the case of
`staked-in needle syringes, once the fluid path is formed by the
`tungsten pin, a needle is securely placed in that fluid path by
`means of an adhesive. During this entire process when the
`tungsten pin is heated to extremely high temperatures, it
`could form tungsten oxide which in turn can potentially
`interact with the glass to form complex tungsten polyanions
`(paratungstenate-A or paratungstenate-B) which could be left
`behind as tungsten-related residue in the fluid path of the
`syringes. These residual tungsten polyanions can interact with
`the protein and lead to further instability issues (3). The
`amount of residual tungsten polyanion left in the syringe
`which is available to interact with the formulation is more in
`the luer-tip syringes compared to the staked-in needle
`syringes because in the case of the latter, the needle and
`adhesive minimize the exposed surface area of the actual fluid
`path to the formulation. Once this fluid path is created, the
`syringe is further processed for siliconization (followed by
`baking if required) and washing. Following this, a needle
`shield (for staked-in needle) or tip cap (for luer slip) is placed
`on the tip end of the syringe. The overall system is then
`sterilized and used for fill-finish activities where once filling is
`completed, a plunger is placed in the syringe and then sent for
`further finishing operations such as visual inspection, etc.
`Based on the above-mentioned process, description, and
`several literature reports ( 4- 6), the critical parts of the pre(cid:173)
`filled syringe system which constitute product contact surface
`are the glass barrel, elastomeric material used for tip cap/
`needle shield and plunger, silicone oil (Si-oil) used for
`siliconization, residual tungsten from boring the liquid path
`into the tip and the needle, and any residual glue in case of
`the staked-in needle syringe. Hence, it is important to
`understand the effect of these contact parts on the stability
`of protein therapeutics (3).
`Although several biotech products such as Enbrel,
`Prevnar, Epogen, Aranesp, Humira, Pegasys, and Peg-Intron
`are available in PFS, common approaches taken to develop
`biotechnology products are not readily available, and there is
`very little published literature available that describe the
`issues of protein product development through key
`approaches and case studies. A few reports on the issues that
`are faced by the industry during the development of protein
`therapeutics in PFS, especially shedding light on silicone oil(cid:173)
`related issues, have recently been reported. Jones et al.
`investigated the propensity of model proteins to aggregate
`in the presence of Si-oil and found that all the proteins tested
`aggregated in the presence of 0.5% Si-oil irrespective of
`molecular weight and isoelectric point of the protein (7).
`Thirumangalathu et al. studied Si-oil induced aggregation of
`an IgGl monoclonal antibody as a function of agitation,
`temperature, pH, ionic strength, and excipient levels such as
`polysorbate 20 and sucrose. Although agitation-induced
`
`aggregation was observed at higher speeds, addition of
`polysorbate 20 inhibited the aggregate formation (8). Over(cid:173)
`cashier et al. described some technical considerations for
`development of protein therapeutics in PFS including selec(cid:173)
`tion of primary components such as plungers and needle
`shields based on syringe functionality experiments (1).
`In this paper, we have provided general guidance on the
`development of biotechnology products in PFS. We have
`systematically evaluated the effect of Si-oil on the compati(cid:173)
`bility of a mAb formulation through spiking studies and
`accelerated temperature stability studies. A study to compare
`two types of syringe barrels differentiated by the siliconiza(cid:173)
`tion process has been presented. Furthermore, a strategy for
`screening primary components has been provided for deter(cid:173)
`mining optimum primary components for PFS systems.
`
`MATERIALS AND METHODS
`
`Materials
`
`An in-house IgG2 mAb solution at 5 or 10 mg/mL in a
`buffer system at pH 6.0 containing surfactant and a dis(cid:173)
`accharide as excipients was used as a model protein
`formulation for this study. Si-oil (Dimithicone, 350 cs) was
`purchased from Dow Corning, Midland, MI, USA. The vials
`and syringes used in the study were made up of type I glass.
`Syringes were sourced from multiple vendors each using
`different processing methodologies to process and siliconize
`the barrels. Elastomeric closure components which can be
`sterilized by autoclaving (plungers) or ETO sterilization (tip
`caps) were used in this study. These were sourced from
`multiple vendors and are free of natural rubber and molded
`from standard formulations that meet requirements of the
`major pharmacopoeias.
`
`Silicone Oil-Water Emulsion Preparation
`
`A stock solution of silicone oil/water emulsion was
`prepared as described earlier by Jones et al. (7). Si-oil was
`added in deionized water in a 50-mL centrifuge tube to obtain
`Si-oil concentration in water of 30 mg/mL. This solution was
`sonicated (Misonix S-4000) for approximately 5 min at 50%
`amplitude to obtain the emulsion which was opalescent due
`to fine dispersed Si-oil droplets. The stock solution was
`diluted to obtain emulsions at concentrations of 10 and
`20 mg/mL Si-oil in water. The freshly prepared solutions
`were used to spike Si-oil into the mAb formulations.
`
`Agitation and Real-Time Stability Study Setup for the Spiked
`Samples
`
`Type 1 glass vials (2 mL) were filled with 1 mL of the
`mAb formulation at 10 mg/mL and were spiked with 0.5, 1.0,
`or 1.5 mg equivalent of Si-oil. A control arm containing vials
`that were not spiked with Si-oil was included. The vials were
`placed in a box in the horizontal orientation to maximize the
`air-water interface. The boxes were then placed on orbital
`shakers at 5°C and room temperature and agitated at 300 rpm
`for 5 and 3 days, respectively. A vial from each group
`(control, 0.5, 1.0, and 1.5 mg/mL Si-oil) was removed after
`days 1, 2, and 3 at room temperature and after days 1, 3, and
`
`Regeneron Exhibit 1044.002
`
`

`

`566
`
`Badkar et al.
`
`5 at 5°C. In addition to the agitation stress study, a separate
`real-time and accelerated temperature stability study at 25°C
`and 40°C was initiated for vials that were spiked with the
`various levels of Si-oil. All stability vials were placed in the
`inverted orientation to maximize contact with the closure as a
`worse-case scenario.
`
`Stability Study Using Pre-filled Syringes
`
`Two types of syringes, each with a different processing
`method, were used. The first type of syringe used was a luer(cid:173)
`tip syringe with baked-on silicone. This syringe was subjected
`to a "baking" process post-siliconization wherein the syringes
`are exposed to a specific temperature for an appropriate time
`resulting in longer chains of Si that are more closely adhered
`to the surfaces they coat, thus resulting in a reduced
`concentration of free silicone in these syringes and lower
`chemical reactivity (9). The second type of syringe evaluated
`was one with the staked-in needle with silicone sprayed-on.
`The mAb formulations were filled into these syringes with a
`nominal fill of 1 mL, and plungers were placed using vacuum
`stoppering resulting in minimal ( <1 mm) headspace between
`the liquid and the plunger. Syringes were then placed on real(cid:173)
`time and accelerated temperature stability at 5°C, 25°C, and
`40°C. All stability syringes were placed in the horizontal
`orientation to maximize contact with both the tip cap as well as
`the plunger stopper. For the purpose of this study, the two
`different types of syringes are referred to as baked-on or sprayed(cid:173)
`on, respectively, in the "RESULTS" and "DISCUSSION"
`sections.
`
`Tip Cap Compatibility
`
`Three different tip cap materials from two vendors were
`used in this study. An accelerated stability study at 25°C was
`initiated to discern the difference between tip cap materials
`and any potential effect on the stability of the mAb.
`Formulations at 5 mg/mL were used for this study. Each tip
`cap was cut into four pieces and added to a vial containing
`1 mL of the mAb formulation. The surface area-to-volume
`ratio of the tip cap area exposed to liquid in the vial was kept
`constant for all three tip caps. A control arm was also added
`to the study where no tip cap material was added to the vials.
`
`Analytical Methods for Analysis
`
`Protein concentration was calculated by using the
`extinction coefficient of the mAb and measuring the UV
`absorbance at 240 nm using a Cary 400 UV spectrophotom(cid:173)
`eter (Agilent Technologies, Santa Clara, CA, USA). For(cid:173)
`mation of soluble aggregates was measured by SEC-HPLC
`using an Agilent 1100 system equipped with G3000SWXL
`and G2000SWXL columns in series. The detection wave(cid:173)
`length was 214 nm using a diode array detector and a nominal
`protein concentration of 1 mg/mL. The protein was eluted
`isocratically with a 200-mM sodium phosphate (pH 7.0)
`mobile phase at a flow rate of 0.7 mL/min. Oxidation was
`assessed using a product-specific reversed-phase HPLC using
`an Agilent 1100 system equipped with two C18 columns in
`series. Enzymatic digestion was used to digest the mAb, and
`the met-256 containing fragment and its oxidized counterpart
`
`were monitored at 214 nm using a diode array detector and a
`nominal protein concentration of 1 mg/mL. All the afore(cid:173)
`mentioned analytical assay methods had been validated
`specifically for the mAb that was studied. Subvisible particle
`analysis was performed using a HIAC 9703 equipped with an
`HRLD-150 sensor. A 15-µm Duke latex count standard was
`run as a system suitability check prior to any sample analysis.
`In addition, a water blank was analyzed before each sample
`to ensure the background counts were appropriate for testing.
`All samples were equilibrated to room temperature before
`analysis.
`
`RESULTS
`
`The aim of this report was to highlight the issues we
`experienced and approaches we took when developing a
`mAb product in PFS presentation. This should also provide
`some recommendations on approaches that can be used for
`development of similar products in PFS.
`
`Silicone Oil Compatibility: Spiking Approach
`
`Si-oil levels typically observed (based on input from
`multiple syringe manufacturers) for staked-in needle
`syringes (0.5 to 0.8 mg/mL) were used for this study as a
`realistic representation of Si-oil. The mAb solution was
`spiked with stock Si-oil emulsion to achieve 0.5, 1.0, and
`1.5 mg/mL silicone oil in mAb formulation as described in
`the "MATERIALS AND METHODS " section. These
`samples were subjected to an agitation stress study and a
`real-time and accelerated temperature stability study. A
`control arm consisting of formulations that were not spiked
`was also included in the study. The results for agitation
`stress study are shown in Fig. la, b for 5°C and 25°C,
`respectively. Soluble aggregate levels as measured by SEC
`did not show any appreciable difference when compared
`with control at 5°C and 25°C for up to 5 and 3 days,
`respectively, at an agitation speed of 300 rpm. Further(cid:173)
`more, no differences in appearance, protein concentration,
`or pH were observed. The stability study results, presented
`in Fig. 2a, b, show no significant change in soluble
`aggregate levels as a function of time and silicone oil
`spiking levels at 5°C. At the accelerated temperature
`condition of 25°C, a slight increase in soluble aggregates
`over 12 months was observed. Varying Si-oil spiking levels
`did not impact rate of formation of soluble aggregates. At
`5°C, there was no observed increase in soluble aggregates
`even after 1 year of storage irrespective of silicone oil
`spiking level. In addition, no changes in color, clarity, pH,
`and protein concentration were observed over the duration
`of the study at either of the two temperatures for all the
`arms including control. All of the Si-oil spiked samples
`stored at 25°C had visible particle formation at the 1-year
`time point. This was also seen for samples spiked at the
`1.5-mg Si-oil level at 5°C at the 1-year time point. No
`visible particle formation was observed for control samples
`at either temperature. Further investigation and particle
`characterization using a combination of optical microscopy
`(polarized light) and Fourier transform infrared spectro(cid:173)
`scopy (FTIR) was conducted. Results from optical micro(cid:173)
`scopy are shown in Fig. 3. The particle morphology and
`
`Regeneron Exhibit 1044.003
`
`

`

`Development of Biotechnology Products in Pre-filled Syringes
`
`567
`
`a
`
`b
`
`5 ~ - - - - - - - - - - - - - - - - - -~
`
`4
`
`3
`
`l
`"' 2
`"'
`"' ~
`"' "' <(
`" 2
`
`1'i
`:,
`0
`(j)
`
`e 1 X Silicone Level
`0
`2X
`! gntrol
`
`0
`
`e
`
`e
`
`0
`
`OL_----'-------'---'-----'-------~---'----__J
`4
`0
`2
`5
`6
`
`0
`
`0
`
`Time (Days)
`Time (Days)
`Fig. 1. Soluble aggregate levels for silicone oil spiked samples when subjected to agitation stress at a S°C and b 2S°C, respectively
`
`reflectivity/opacity properties under varying light condi(cid:173)
`tions suggest that the particles seen for samples at 25°C are
`a mixture of Si-oil droplets along with very small amor(cid:173)
`phous particles that resemble protein particulate matter.
`The shape and morphology of the silicone oil droplet
`observed was similar to that reported by Narhi et al. (10).
`Further characterization of these particles using FTIR
`spectroscopy confirmed that the amorphous particles
`observed were proteinaceous in nature. It can be hypothe(cid:173)
`sized that, in the presence of Si-oil, prolonged exposure to
`accelerated temperature conditions has the potential to
`cause small amounts of protein to precipitate on the Si-oil
`droplets resulting in visible particles containing Si-oil and
`protein matter (Si-protein particle). It is important to note
`that the formation of these visible particles did not have
`significant impact on either protein concentration or
`soluble aggregate formation, indicating that the amount of
`protein precipitating was very small and within the assay
`variation for the concentration assay and that the forma(cid:173)
`tion of soluble aggregates was independent of the amount
`of Si-oil in the sample. Particle characterization of samples
`stored at 5°C, which had visible particle formation at the 1-
`year time point, indicated that they were Si-oil droplets
`without the presence of any protein-related matter. This
`suggests that samples containing high levels of Si-oil may
`be challenging for visual inspection due to the potential for
`
`a
`
`the Si-oil to coalesce over time resulting in higher levels of
`rejects of the drug product.
`
`Selection of Primary Components: Type of Syringe Barrel
`and Siliconization Process
`
`The selection of primary components such as syringe
`barrel, plunger, and tip cap and/or needle shield is an
`important decision from a container closure standpoint.
`Oftentimes, several options are available, and it is important
`to choose the right combination of primary packaging
`material that can provide adequate stability for the molecule
`of interest as well as maintain the desired functionality
`properties over the product shelf life. This section highlights
`a study we conducted for choosing the correct type of syringe
`barrel for a mAb product. Two types of syringes, each with a
`different processing method, were used as described in the
`"METHODS AND MATERIALS" section. Real-time and
`accelerated temperature studies of the model mAb formula(cid:173)
`tion in PFS were initiated. Based on extraction studies
`performed in our labs ( data not shown), it was determined
`that the sprayed-on syringes contained higher residual-free
`silicone compared to baked-on silicone syringes. Figure 4a-c
`depicts soluble aggregate levels obtained for the 10-mg/mL
`mAb formulation stored in either sprayed-on or baked-on
`syringes at 5°C, 25°C, or 40°C, respectively. No specific trends
`
`b
`5~ - - - - - - - - - - - - - - - - - -~
`
`(t)
`
`~
`"'
`1ii
`"'
`~
`"' "' <(
`
`(t)
`1'i
`:,
`0
`(j)
`
`4
`
`2
`
`I
`
`0
`
`• 1 X Silicone Level I
`..
`"' Control
`
`2 X Silicone Level
`3 X Silicone Level
`
`1 ~ Q
`
`Q
`
`Q
`
`Q
`
`8
`
`0
`0
`
`2
`
`4
`
`6
`
`10
`
`12
`
`14
`
`16
`
`18
`
`20
`
`4
`
`2
`
`l
`"' 2
`"' 0)
`~
`0)
`"' <(
`
`Q)
`]5
`:,
`0
`(j)
`
`~
`
`0
`0
`
`8
`
`!:,
`
`l'i
`
`"'
`l'i
`
`2
`
`4
`
`6
`
`8
`
`10
`
`12
`
`14
`
`Time (months)
`Fig. 2. Soluble aggregate levels for silicone oil spiked samples at a S°C and b 2S°C, respectively, when subjected to real-time
`and accelerated temperature conditions, as a function of time
`
`Regeneron Exhibit 1044.004
`
`

`

`568
`
`Badkar et al.
`
`3X-5C Sample
`
`2X-25C Sample
`
`0 II'
`
`I
`I
`I
`I
`I
`I
`\
`I
`
`/
`
`/
`
`/
`
`/
`
`/
`
`/
`
`/
`
`/
`
`/
`
`,4
`/
`
`/•
`
`/
`
`/
`
`-J.f
`
`0 0
`
`,t,
`
`/
`
`/
`
`/
`
`/
`
`I
`
`I
`
`I
`
`/
`
`I
`
`I
`
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I/
`
`/
`
`/
`
`/
`
`/
`
`\
`I
`/
`/
`I
`Silicone Oil Droplets
`Fig. 3. Optical microscopy photographs showing silicone oil droplets and amorphous proteinaceous
`particles for 25°C sample spiked with 1 mg silicone oil and only silicone oil droplets for 5°C samples
`spiked with 1.5 mg silicone oil spiking
`
`Amorphous proteinaceous particles
`
`were observed in the stability profiles for both syringe types
`in the case of samples that were stored at 5°C, although a
`slight increase in soluble aggregate level was observed for the
`sprayed-on syringes stored at the accelerated temperature
`condition of 25°C. At 40°C, a significant increase in soluble
`aggregate formation was observed for the sprayed-on
`syringes. All the samples were visually free of particulate
`matter. In addition to this, the samples were evaluated for
`subvisible particle counts per USP <788> at both 5°C and 25°
`
`C. For this assay, in addition to the USP-specified channels of
`10 and 25 µm, particle counts at lower size-range channels
`such as 2, 5, and 8 µm were also measured to gather
`additional information in order to determine trends. As
`expected, the results from the subvisible particulate analysis
`on samples stored in baked-on syringes were within the USP
`<788> limits for 10 and 25 µm, and no specific trends were
`observed at the lower size channels as a function of time
`(Fig. Sa). Subvisible particle count levels for sprayed-on
`
`b
`
`4
`~
`
`"' * 3
`
`Ol
`~
`Ol
`Ol
`<(
`Q) 2
`J5
`::,
`0
`(/)
`
`1,
`
`0
`
`0
`
`e Staked-in Needle I
`0 Luer Cone
`
`I
`
`e
`
`~
`
`•
`
`4
`
`10
`
`12
`
`14
`
`Time (months)
`
`a
`
`e Staked-in Needle I
`0
`Luer Cone
`
`I
`
`• 0
`
`• 0
`
`• 0
`
`• 0
`
`•
`
`0 ~~~~~~~~~~~~~~~~~~~~~
`10
`12
`0
`6
`14
`
`Time (months)
`
`C
`s~ - - - - - - - - - - - - - - - - - - -~
`
`• Staked-in Needle I
`0
`Luer Cone
`
`I
`
`•
`
`•
`
`0
`
`)
`
`o~~~~~~~~~~~~~~~~~~~~~
`0
`
`Time (months)
`Fig. 4. Soluble aggregate levels for spayed-on and baked-on syringes at a 5°C, b 25°C, and c 40°C as a function of time
`
`Regeneron Exhibit 1044.005
`
`

`

`Development of Biotechnology Products in Pre-filled Syringes
`a
`
`1600
`
`b
`
`12000
`
`1400
`
`569
`
`0
`
`:. C 1200
`'iii
`c
`~ 1000
`.,
`:l:l
`800
`t: " Q.
`0
`:. .0
`E
`:, z
`
`600
`
`400
`
`200
`
`0
`
`-
`
`-
`
`lnffial
`6month
`12month
`
`:;; 10000
`C
`'jij
`c
`
`6000
`
`0 u '" "' :g
`
`iii
`0..
`0
`
`'"
`.0
`§
`z
`
`6000
`
`4000
`
`2000
`
`2um
`
`Sum
`
`Sum
`
`101MT1
`
`25 um
`
`2um
`
`.S urn
`
`Bum
`
`10um
`
`25um
`
`Particle Size
`Particle Size
`Fig. 5. Subvisible particle levels for a baked-on and b sprayed-on syringes (respectively) at 5°C as a function of time
`
`syringes were significantly different than those seen for the
`baked-on syringes especially for the lower particle channels
`(Fig. Sb) with an increase in subvisible particles for size
`channels 2, 5, and 8 µm observed for the sprayed-on syringes.
`It is also worthwhile to note that particle counts for samples
`analyzed at the 12-month time point stored at 5°C and 3- and
`6-month time points stored at 25°C (data not shown) were so
`high that they saturated the detector, and no measurements
`could be made for those samples. Further particle character(cid:173)
`ization using optical microscopy indicated that the particles
`were identical to those seen in the Si-oil spiking study, and the
`shape and morphology of the particles was correlated very
`nicely with that reported by Narhi et al. (10) and was
`therefore attributed to small Si-oil droplets (Fig. 6). Overall
`data suggested that the baked-on silicone process was better
`suited for protein formulation development in PFS as it
`represented a lesser degree of risk for the formation of
`subvisible particulate matter as well as minimized any
`potential for protein precipitation on the Si-oil droplets.
`
`Selection of Primary Components: Tip Cap Compatibility
`
`In this section, we have provided an example of an
`approach that was used for choosing the correct type of tip
`cap for development of a mAb product in PFS system.
`Although tip cap material is considered to have minimal
`product contact, it is imperative that optimum material is
`chosen. Three different tip cap materials from two vendors
`were evaluated against the mAb formulation. Tip cap samples
`
`r\-::: __ u ------
`0 _ ----------~~----:._- Si Oil Droplets
`~~---
`~ -"-----------
`
`~-
`
`Fig. 6. Optical microscopy image showing that the subvisible particles
`obtained from sprayed-on silicone syringes
`
`were cut and placed in vials containing mAb formulation as
`described in the "MATERIALS AND METHODS" section.
`These samples were evaluated for changes in quality attrib(cid:173)
`utes over 6 months when stored at 5°C and 25°C. A control
`arm with no tip cap in the formulation was also placed on
`stability. Figure 7a, b shows no substantial difference in
`soluble aggregate level between the different tip cap materials
`as compared to the control at both temperatures, with all the
`materials fairing equally. Figure Sa, b shows oxidation level
`changes as a function of time and tip cap material at 5°C and
`25°C, respectively. Increase in oxidation as a function of time
`was observed for both 5°C and 25°C samples over 6 months
`for materials A and B as compared to the control. Material C
`did not show any significant increase in oxidation over the
`same time period when compared with the control samples.
`This observation was significant and prompted the choice of
`material C for the tip cap.
`
`DISCUSSION
`
`In this paper, we have limited our investigation to two
`broad aspects: stability of a model mAb formulation in the
`presence of practically relevant levels of Si-oil and a stability(cid:173)
`based approach for selection of primary packaging compo(cid:173)
`nents (syringe barrel and tip cap). Discussions on the
`selection of the plunger stopper were not in scope of this
`study since the elastomer resin selected for the plunger
`stopper for this mAb product was identical in composition
`to that of the vial stopper for the mAb product in vials.
`Hence, no additional compatibility work was required prior
`to the stopper selection. However, an approach similar to the
`one taken for the tip can also be used for the plunger stopper.
`
`Silicone Oil Compatibility
`
`Silicone oil is an essential component of PFS due to its
`role as lubricant to achieve desired functionality ( e.g.,
`maintain a certain break-loose force and glide force over
`shelf life). Proteins are complex molecules and due to their
`amphiphilic nature can interact with a variety of surfaces.
`These interactions can result in structural changes in the
`protein leading to denaturation. Presence of a silicone oil
`layer as the main contact surface in PFS systems can have
`adverse effects on protein stability (7). Under typical
`
`Regeneron Exhibit 1044.006
`
`

`

`570
`
`a
`
`5
`
`4
`
`""
`
`!c_..,
`<I)
`
`* 3
`
`a,
`~
`a,
`a,
`<(
`"' 2
`:0 ::,
`0
`(/)
`
`• Control
`0
`! Tip Cap B
`
`Tip Cap A
`
`Tip Cap C
`
`0
`
`0
`
`0
`
`0
`
`b
`
`5
`
`Badkar et al.
`
`e Control
`0
`Tip Cap A
`! :i::~ g:~~
`
`4
`~ !c_..,
`
`3
`
`<I) * a,
`
`~
`a,
`a,
`<(
`
`"' :0
`::,
`0
`(/)
`
`2
`
`e 0
`
`0
`
`0
`
`2
`
`4
`
`5
`
`6
`
`3
`Time (months)
`Time (months)
`Fig. 7. Soluble aggregate levels for mAb formulation samples at a S°C and b 2S°C (respectively) when in contact with
`various tip cap materials
`
`0
`
`0
`
`2
`
`conditions, protein solutions in pre-filled syringes can be
`exposed to several types of stress factors (shaking during
`shipping, temperature excursions, etc.) while in contact with
`silicone oil-water interface over its shelf life. Therefore, it is
`imperative that the compatibility of protein formulation with
`silicone oil be assessed for the development of protein
`formulations in PFS. In this paper, we have presented an
`approach for evaluating the compatibility of protein formu(cid:173)
`lation with silicone oil. A silicone oil/water emulsion was
`prepared and used to spike the required amount of silicone to
`the mAb formulation in vials to represent a scenario of
`silicone oil level which is typically seen in PFS. The silicone
`oil level can vary depending on the syringe manufacturer,
`syringe size, and processing conditions. A survey of leading
`PFS manufactures resulted in our finding that the typical
`levels reported vary between 0.5 and 1 mg silicone per
`syringe. Hence, we selected 0.5, 1, and 1.5 mg as the varying
`levels of silicone oil for our spiking study. Once the samples
`were spiked with the determined amounts of silicone, our
`objective was to simulate and/or exaggerate the types of stress
`that a typical syringe would experience during its lifetime on
`the shelf. Agitation and temperature were identified as the
`major stress factors to be evaluated. Results from the
`agitation stress study suggest that no deleterious impact on
`any of the quality attributes monitored was seen when
`
`samples were agitated with speeds of up to 300 rpm. In
`particular, there was no increase in soluble aggregates and no
`changes in appearance of the solution during the entire
`agitation study duration of up to 5 days at 5°C for all the
`samples including control. Placing vials in the horizontal
`orientation during the agitation study helped maximize the
`air-water interfacial area available for protein adsorption.
`The results follow similar trends as those seen with the
`previous study presented by Thirumangalathu et al. , where
`authors observed that aggregation of the monoclonal anti(cid:173)
`body was promoted at agitation speeds of 2350 rpm.
`Interestingly, the authors also reported that there was no
`aggregation seen at agitation speeds of less than or equal to
`200 rpm (8). The agitation speed of 300 rpm used in this study
`is quite vigorous and, from our visual observations, thought to
`be equivalent to (if not more rigorous) expected agitation
`experienced during shipment or under recommended storage
`and shipping conditions.
`Although no deleterious effect on protein qu